Sensor Based Detection of Implant Loosening in Total Hip Replacements

Final Report Summary - SMART-HIP (Sensor Based Detection of Implant Loosening in Total Hip Replacements)

Executive Summary:
Ageing, sedentary behavior, and obesity are predictors of osteoarthritis, a Non Communicable Disease which is the leading cause of disability in the developed world, affecting 9.6% of men and 18% of women aged over 60 years, and among the leading conditions causing work limitations. (WHO, 2010a). Moreover, the combination of arthritis and obesity causes an undue stress on joints, and when conservative treatments no longer yield a satisfactory response, joint replacement becomes necessary to relieve pain and stiffness and improve quality of life.
Despite the standardized procedures, a significant share of total joint replacements fail because the prosthesis becomes loose or because of osteolysis, and the prosthesis must be replaced. Bone loosening is usually diagnosed by radiography and clinical symptoms, but pre-operative radiographic diagnosis of loosening has a sensitivity of 80%, and a considerable number of revision surgeries is not necessary because loosening of the total joint replacement was diagnosed false positively.
Currently clinically applied methods of assessing implant fixation and implant loosening are of sub-optimal precision, leading to unsecure indication of revision surgery and late recognition of bone defects.
To solve this technology gap, the SMART-HIP goal is to develop a new intelligent hip prosthesis, enabling timely and accurately diagnosis of bone loosening, thus allowing for a fast and reliable support to the orthopedists while deciding upon revision surgery of joint replacements.
The final SMART-HIP system will be composed of two fundamental sub-systems:
The Intelligent Prosthesis, whose basic component will be the Oscillator Unit consisting of a magnetic or magnetisable body which is fixed on a flat steel spring, which will enable to perform an acoustic-mechanical analysis of the variation of the resonance frequencies of total joint replacements, correlating them with the loosening status of the bone.
The Diagnostic Device, which is placed outside the patient’s body and used during the clinical test to excite the oscillator. This excitation within the implant bending modes leads to a sound emission to the surrounding bone and soft tissue, which can be detected by a vibration sensor, which is applied outside the patient.
Successful realisation of the innovative SMART-HIP system will enable accurate and reliable diagnosis of implant loosening. Besides the obvious benefits to osteoarthritic patients, commercialisation will enhance the competitiveness of partnering small businesses while reducing disability-related socioeconomic costs.

Project Context and Objectives:
Project Objectives
As reported in the Description of Work, the final SMART-HIP system was meant as composed of two fundamental sub-systems:
1. The SMART-HIP Intelligent Prosthesis (EIP): basic component of this part of the system is represented by the Oscillator Unit, seamlessly integrated in the EIP, consisting of a magnetic or magnetisable body which is fixed on a flat steel spring. This oscillator enables to perform an acoustic-mechanical analysis of the variation of the resonance frequencies of total joint replacements, correlating them with the loosening status of the bone.
2. The SMART-HIP Diagnostic Device (EDD): the proposed system for clinical use consists of the following sub-systems:
The External Coil, placed outside the patient’s body and used during the clinical test to excite the oscillator. The oscillator itself impinges on the joint replacement and thereby creates a sound signal (implant resonance frequencies).

The Detection Unit: the excitation within the implant bending modes leads to a sound emission to the surrounding bone and soft tissue, which can be detected by a vibration sensor, which is applied outside the patient. To this scope, an adequate sensor device must be developed in order to measure the produced sound signal. A frequency range of at least 0 28 kHz shall be measurable with a sample rate of at least 150 kHz. The sensor device must be designed such that it is insensitive to manual application, e.g. when pressed on the patient’s skin.

In addition, for analysing the measured frequencies, custom designed software is needed. The software includes automatic Fast Fourier Transformation (FFT) as well as determination of the centre frequency of the normalized spectrum. From these values, bony integration of the implant is evaluated.
The key milestones for the realization of the SMART-HIP system are summarized below:
1. Mechanical design of the Excitation Unit to meet the Regulatory Constraints;
2. Excitation Unit able to withstand the sterilization process;
3. Design of an Excitation Unit able to create sound emission that can be measured from the patient’s skin surface;
4. Choice of a sensor to be included in the measurement device;
5. Development of hardware: extracorporeal coil and Evaluation Unit;
6. Coding of software with semi-automatic evaluation of measured sound signals;
7. Successful test of SMART-HIP system in a cadaveric study.
The key challenges to overcome in the project are:
1. Producing a measurable sound signal and keeping the excitation unit as small as possible
2. MRI compatibility
3. Evaluation of small frequency changes
4. Development of an optimized filtering algorithm to exclude influence from body movements, skin elasticity
5. Determination of the sensitivity in cadaveric experiments: simulation of loosening conditions

The SMEs in the project have immediately started working on the definition of the main system requirements and initial specifications, which have been thought to fit the current needs of the orthopedic industry.
To further focus the product development, WP1 intended to ensure that the system fits all the regulatory and medical standards and customers’ needs. Besides, we have conceived this WP so that a series of preliminary concepts fulfilling all the requirements were prototyped (through CAD design or rapid prototyping facilities), so as to give the SMEs concrete elements to decide upon the concept design to be pursued.
Then, the project focused on the development of the two main project results:
The SMART-HIP Oscillator Unit, integrated in the prosthesis;
The SMART-HIP Diagnostic Unit.
Although requiring two separate research lines, their design and development has been conceived to be carried out in a single WP (WP2 – Hardware Development), in order to be sure that the overall system designed is coordinated and handled through a single task force.
In parallel with this, Software Development was performed in WP3, with the design and implementation of the algorithms to process the data received by the Diagnostic Unit, as well as the design of the User Interface.
According to this structure, the two paths (WP2 and WP3) could be followed quite independently, and converge in the final system integration (WP4), where the pre-industrial release of the SMART-HIP system was prepared for the final tests and Medical Validation, performed under WP5.
Work performed in the project
The following section provides a brief description, at a WP level, of the main activities performed during the whole project (M1-M24). The detail of each WP’s results will be reported later on under the “S&T foregrounds” section.
Work performed under WP1 - CONCEPT DEVELOPMENT
WP1 was mainly devoted to the identification of the main application scenario for the SMART-HIP project, to research and analyze the markets of interest together with the regulatory field as well as to identify requirements and specifications for all the components, which represent the basis for the product technical specifications.
The analysis of the application scenario identified, Total Hip Replacements, shows that, in the total hip replacement framework, there is an urgent need of the determination of implant loosening in detection method but, currently, the imaging-based diagnostics used are not specific for the problem. The result is that the need remains currently unmet.
All SMEs, led by the WP leader MERETE, have conducted workplace studies and survey in the target application market supported also by the RTDs partners. A market study was carried out comprising the occurrence of inserted hip stems and acetabular cups in primary surgery. The market research includes data from the European framework and, then, the most interesting market segments have been investigated: Germany and Sweden.
The analysis of the state of art of the most common techniques currently in use for the implant loosening detection shows that there is a lack of specificity and, when the specificity is high enough, the technique is far to be a standard in the clinical practice due to the high costs associated.
Focusing on the use application scenario identified, a Requirements/Specification matrix has been implemented in order to map all the possible system configurations that can lead to a successful concept development. Therefore, during the three months of WP1 a preliminary concept of the system architecture of the Smart-hip project has been depicted, also in accordance with the involved European regulations regarding medical devices.
In addition, the integration hypothesis of each of the parts mentioned above have presented within this WP. It was a preliminary idea mainly regarding the possibility of merging together the Excitation Coil and the Detection Unit, which will be further developed during the second period of the project within the WP4 – System integration, working on optimization and taking into account the final user’s needs.
Finally, the Ethical Manual of the SMART-HIP project has been implemented, in which all the provisions necessary to duly collect, process, store and dispose of the anonymized samples that will be used in the project are described. Ethical Clearance was achieved from the Ethics Committee, University of Rostock, from April 16th, 2014 under Reg.# A 2014-0055.
Work performed under WP2 – HARDWARE DEVELOPMENT
This WP started in month 4 and continued until month 13. The main goals within the WP2 were intended to be:
1.Hardware design activities of the SMART-HIP system;
2.Components prototyping;
The design activities have been carried out taking into account the analysis developed in the WP1, by implementing the architecture already presented and confirmed by the Consortium.
The first phase of this work package was devoted to system design activities over the following four main blocks:
1) Oscillator Unit
2) Excitation Coil
3) Detection Unit
4) Personal Computer
The blocks 1 and 2 are responsible for sound generation, crucial for the loosening detection while the blocks 3 and 4 are the main components of the Diagnostic Unit, where the sound is detected, computed and shown to the final user by means of an HMI (Human Machine Interface, developed within WP3).
The Oscillator Unit, composed of a metal ball enclosed in a screw installed in the upper part of the prosthesis – which generates a sound signal proportional to loosening – was designed and functional samples were produced. During the Period 2 the Oscillator Unit was transferred to a prototype.
The Excitation Coil controls the movement of the metal ball. First preliminary functional tests have been performed and measurements could be recorded. Several layouts of Excitation Coils were designed and built. From these different layouts of Excitation Coils, a prototype has been developed and transferred to a prototype.
Although not mentioned in the DoW, a Coil driver was strongly needed in order to replace the bench power supply, which could not be integrated with the Detection Unit. Without such Driver, the integration activities would have been very difficult and the final system not completely exploitable as a custom product. Therefore, in order to fix that issue, LABOR gathered the main electrical specifications (waveform, amplitude, power etc.) and designed an integrated unit, to be used along with the Detection Unit and directly controlled by the latter.
The Detection Unit is an electronic device, composed of
1.An analog stage including an IEPE piezoelectric accelerometer (responsible for the sound detection)
2.An analog to Digital stage
3.A digital microcontroller unit, which transfers measurements to the Personal Computer, where the HMI runs.
The next steps of the design activities included further preliminary tests and improvements, along the path already traced. Right afterward the prototyping phase started: prototypes of the mentioned components were produced and preliminarily tested by both UMR and LABOR. The experimental results so obtained, resulting from each single part finally developed and built, have been reported in the D2.2 – System Mock-up.
Work performed under WP3 – SOFTWARE DEVELOPMENT
This work package ran in parallel with WP2, its work being completed at month 13.
DIAGNOSTIC SOFTWARE:
UMR set up a test bench with flexible hardware, capable of generating and acquiring the signal from the sensors after excitation.
Afterward, the first phase is to design the software acquisition with precise features of acquire and display the signal. Also the capability of the diagnostic software is to save on the file with a TDMS format, similar to a spread sheet file divided per categories.
The second phase is to design the analysis software to load the files previously stored according to the specification above and to implement analysis. This is achieved by a pre-processing, filtering and frequency domain transformation.
After the first analysis results, data are ready to be introduced to loosening analysis. The first method implements a peak detection algorithm. The second method considers the spectrum with linear amplitude. This algorithm is also currently available as a MATLAB script.
At the end of the first period, a time domain analysis approach was presented as attenuation analysis of the time signal: in this analysis approach an approximated envelope curve has to be fitted into the fading time signal. The parameters of this curve have to be analyzed regarding the decay properties and the damping coefficient.
HMI:
The HMI is designed as a standard software available to doctors. Therefore, the software designer has to pay attention that the user is assisted during the usage, by providing quick access to the information, quick evaluation of the priorities, clear feedback from the system, compatibility between presented data and expected data, clear information display, information completeness and high level of use, full control of the system, predictability in use, information exchange between two operators.
Regarding the doctor point of view, we implemented a survey questionnaire to be submitted to doctors. For the project purposes, however, we proceeded with standard professional medical software models.
In addition, LABOR gave priority to robust hardware connection to guarantee the signal command, and in a second moment focused on signal results retrieving. In the first release of the system, the PC hosting the HMI is separated from the electronic parts, so the physical connection is made manually before or after software starting by two micro-USB cables, one for the data streaming port, one for the command port. The rest of the system, belt, coils, sensors, will be responsive to these two connection channels.
To integrate the HMI with the most important part of loosening detection LABOR left this part ‘dummy’ intending it as a dummy data produced by simple algorithm. In Period 2 of the project these parts could be easily integrated thanks to common programming and prototyping language: LABVIEW. This is a graphical language thanks to which the diagnostic and analysis software was conceived; the final software could be assembled reusing routines and software loop structures already implemented.
Work performed under WP4 – SYSTEM INTEGRATION
This WP started in month 14 and continued until month. The present chapter will summarize the main results achieved during the related tasks (T4.1 T4.2 T4.3).
The main goals within the WP4 were intended to be:
1.System integration and mock-up: the components already produced during WP2 (Oscillator Unit, Excitation Coil, Detection Unit) were integrated into an overall working mock-up and then preliminary tested;
2.System debug: bugs found during preliminary tests were recorded and studied;
The overall complete system was, therefore, ready to be put in a test environment, conceived to be as real as possible. The accelerometers (connected to the Detection Unit) were placed over the prosthesis (containing the Oscillator Unit), previously inserted into a synthetic bone. The Excitation Coil excited the ball enclosed in the Detection Unit, therefore, the whole system could be tested.
The results so obtained have been reported in the D4.1 – SMART-HIP Prototype.
Work performed under WP5 – FIELD TESTS
The work package 5 started from month 6 and ran until the end of the SMART-HIP project. WP 5 was subdivided into four tasks: Develop testing protocols for the field tests (Task 5.1) run in vitro tests on artificial bone specimen and human femoral specimen (Task 5.2) validate components of the SMART-HIP diagnostic system (Task 5.3) and finally develop an economic analysis (Task 5.4).
The testing protocols for the field tests are described in D5.1. Herein, the assembly of the experimental test setup as well the connections of the electronic components are explained. Moreover, the testing protocol regulates which adjustments are required in the measurement parameters and how the proper documentation of tests needs to be fulfilled.
The field tests which are described in detail in D5.2 started with a rather simplified model using a biomechanical foam cylinder as bone substitute. Three different levels of implant loosening were simulated with the non-cemented Z-Stem implanted in these cylinders. For performing the field tests, all components of the SMART-HIP system were used at their current status of development. During the first months of work package 5, the artificial bone cylinders were tested under dry conditions, i.e. the sensors were placed directly at the cylinder surface. In a subsequent step, the tests were performed under wet conditions, i.e. a water-filled latex tube was integrated in the test setup in order to simulate skin and soft tissue environment around the bone. Thereby, the sensors were placed on the latex tube as planned for clinical application on the patient’s skin. The final step in bringing the field tests closer to the real conditions was using real human bones instead of the biomechanical foam cylinders. Ethical clearance for using human specimen was gained in D1.3. Within the course of the project, updated hardware and software was used at the different stages of in vitro testing depending on the experiences made.
Therefore, D5.2 includes a standard operating procedure on how to perform testing with human specimen under the latest level of development of the SMART-HIP system.
Data analysis within the field tests was performed using principal component analysis and support vector machines. Using this technique, an almost 100% accuracy in detecting the different levels of implant loosening was achieved. In order to be able to compare the acoustic measurements to mechanical parameters, pull-out tests were performed on the human femur specimen extracting the endoprosthetic stem from the bone bed. Significantly lower pull-out forces were measured with larger level of loosening.
The economic assessment as described in D5.3 addresses the main parts of the SMART-HIP system, which is composed of the Diagnostic Device (Extracorporeal coil, Detection Unit, Coils driver), Software and HMI as components of the extracorporeal Diagnostic Device, as well as Oscillator Unit as the implantable device, part of the Intelligent Prosthesis. All factors from manufacturing process, cost analysis and possible pricing are discussed in D5.3 as result of Task 5.4.
A specific task was foreseen in this WP aiming at the creation of a feasibility study of the SMART-HIP System, composed of the SMART-HIP Diagnostic Device, the SMART-HIP Software and the SMART-HIP Intelligent Prosthesis (hip stem + oscillator unit).
Work performed under WP6 – DISSEMINATION, EXPLOITATION AND MARKETING
The WP6 ran for the entire project lifecycle and the focus of the activities performed within this work package were addressed to the full utilization of the results to obtain system's introduction during and after the end of the project. The main objectives achieved were:
1. The maintenance of SMART-HIP website, which was updated periodically during the whole project;
2. The update of the main dissemination tools to be used for promotional purposes;
3. Creation of a final press release on the results of the project;
4. Continuous diffusion of the project's awareness through the main social networks (Facebook, Twitter and LinkedIn) as well as through the project link from the SMEs’ official pages;
5.Implementation of the dissemination strategy concerning the list of activities planned and carried out during the Period 2 (M10-M24) created and validated by the Consortium to properly coordinate the communication and promotional activities to be carried out during the second phase of project;
6.Development of the SMART-HIP exploitation plan, characterizing the exploitable results of the project and giving information about IPRs background and foreground as well as the exploitation claims.

Project Results:
Overview
The activities carried out in the project (M1 – M24) were mostly devoted to the achievement of the following strategic goals, from a technical point of view:
1.Design and implementation of the SMART HIP hardware, meaning the manufacturing and assembling of the oscillator unit, of the diagnostic unit and of the external coil (WP2);
2.Design and implementation of the SMART HIP software, activity started in P1, which include the signal processing algorithm definition, the diagnostic software and the human machine interface development, which constitute the final SMART-HIP software (WP3);
3.Integrating the specific modules developed in the previous WPs in the SMART-HIP device, performing preliminary tests on the prototype to verify is proper functioning and debugging the system also evaluating possible adjustments to both the hardware and the software, if needed (WP4);
4.Defining the test protocol to be used for the validation of the SMART-HIP device and performing an extensive in vitro testing phase, followed by measurements on real bones to verify the effective response of the diagnostic device and the precision of the output data (WP5);
5.Performing an economic analysis and an evaluation of the cost-effectiveness of the developed solution, providing an overview of the future incomes coming from the SMART-HIP device commercialization, also conducting an analysis of the state-of-the-art and of the competitors of the new technology (WP5);
6.Carrying out intense dissemination and communication actions to widely diffuse information and results of the project to external audience, stakeholders and end-users, and planning exploitation for the future SMART-HIP product taking into account the specific interests of each SME in the Consortium (WP6);
7.Performing the Management of the Consortium, quality control tasks and supervision of the overall workplan development by the partners, contributing to the creation of the periodic and final reports for the project (WP7).
We can summarize here the main goals achieved in the project:
1.Design and prototyping of the SMART-HIP system components;
2.Implementation of the diagnostic software, of the signal processing algorithm and of the HMI;
3.Definition of the standard protocols for testing the SMART-HIP system;
4.Integration of the developed components in the final SMART-HIP device for subsequent tests;
5.Preliminary tests and debugging of the system, in vitro tests and intense testing phase on real human bones to validate the signal acquisition, verify reliability and reproducibility of measurements, as well as the precision of the acquired data;
6.Economic assessment and cost-benefit analysis, evaluation of the commercial potential of the SMART-HIP solution and of its most appropriate/profitable applications;
7.Creation and update of the project website and to communication tools developed in P1, according to the results obtained, full characterization of the plan for disseminating and exploiting the project after the end of the project.
From the point of view of the Management activities, the following actions have been performed by EDE with the collaboration of all partners:
1.The preparation and submission of the First and Second Periodic Report for the project to the REA, the collection and check of the partners’ financial statements (Form C) for P1 and P2;
2.The creation of a Final Report showing all the developments carried out within the project framework;
3.The preparation of an agreed version of the Consortium Agreement, signed by all partners, and its constant update (new version created for Amendment No. 1 to the GA );
4.The submission of all the planned Deliverables for the project, and the supervision on their technical contents, on the base of what was originally planned in the DoW;
5.The planning and organization of general Project Meetings with all the partners, and the verification of the technical developments obtained by the RTDs in accordance to what is foreseen from the funding scheme, the management and coordination of internal technical meetings at WP level in order to monitor the status of the activities and to validate the results achieved;
5.The application of revision procedures for the official documents and quality control/monitoring actions aimed at reaching the project goals and eventually recovering delays /redirecting the activities towards the final objectives;
6.The preparation of official documents for the Amendment request for the access of a new participant to the Consortium, and the interaction with the Commission services whenever this was needed;
The management of the overall communication flow during the project lifetime, both internal to the Consortium and with the EU Commission, and the handling of the relations between the partners to encourage collaborative atmosphere and fruitful work conditions during the whole project lifetime.
A brief summary of the results per WP is reported below:
WP1- Analysis of the application scenario showed there is a significant need for increased accuracy in loosening diagnostics. To simplify high demands of medical product certification, it was drawn that the hip stem chosen by MERETE should not be changed, instead the Oscillation Unit should be designed as an additional product to be combined with the existing hip stem. A specifically tailored product concept was developed. The ethical clearance for the field tests using human bone was achieved.
WP2- All the system components were designed and manufactured on the base of the requirements and specifications gathered in the course of WP1; these were preliminarily tested by LABOR and UMR. The SMART HIP oscillation unit, the excitation coil and the detection units were then designed and implemented; in addition, a coil driver was developed and included in the system to replace bench instrumentation and avoid integration problems with the detection unit.
WP3- The Diagnostic Software was built in order to realize on the one hand generation of excitation signal for the coil, and on the other hand acquisition of sensor signals, visualization of signals (for standalone usage without HMI unit), recording and storing of measured values as well as data analysis of measured values. Besides that, computational simulations of the implant-bone-compound were conducted in order to estimate the expected frequencies at different loosening levels.
WP4- The Diagnostic Unit developed and built by LABOR in this WP was sent to UMR to integrate it into the test setup including the Excitation Coils and the Oscillation Unit. After a debugging period and some adjustments the system was made operative and measurements could be properly conducted.
WP5- All SMART-HIP components were assembled in a specially designed test setup in order to run field tests on artificial biomechanical bone cylinders and human femur specimen in an environment that mimics the human upper leg. Different loosening levels of an implanted hip stem could be identified with the SMART-HIP diagnostic system with an accuracy of nearly 100% using principal component analysis and support vector machines. Economic assessment was performed from manufacturing processes to first assumptions of pricing.
WP6- Proper dissemination and exploitation activities were planned and conducted under WP6 through the whole project. The application scenarios for the SMART HIP solution, a market segmentation analysis and end-users/stakeholders’ overview was performed. The details on the exploitable knowledge and the results of the Consortium’s intentions for the commercialization of the device were collected in the Final PUDK, where communication activities were also summarized.
WP7- The effective communication has been established between partners and towards the commission. All objectives and milestones have been reached. Four consortium meetings have been organized with all required agendas, minutes and action plans prepared on-time. IPR issues were also handled through this work package.
We can state that:
The intended target system performance and the requirements/specifications matrix was made available (REQUIREMENTS AND SPECIFICATIONS MATRIX) MS1 achieved in P1;
The final designs of the oscillation unit, of the external coil and the diagnostic unit were delivered (SYSTEM DESIGN) MS2 achieved in P1;
Validation of the electronic prototype performed, assessment against the specifications performed (SMART HIP HARDWARE) MS3 achieved in P2;
Assessment of the software performance against the initial specifications performed (SMART HIP SOFTWARE) MS4 achieved in P2;
Final SMART HIP prototype made available (SMART HIP PROTOTYPE) MS5 achieved in P2;
Clinical validation and assessment of the SMART HIP performance respect to state of the art devices fulfilled (CLINICAL VALIDATION OF THE SMART HIP) MS6 achieved in P2.
The Consortium Management has been carried out by EDE during the whole project, and support on the management tasks in the moderation of technical discussions and the preparation of dedicated progress meetings came from the Technical Manager and from the performers.
In summary, we can state that:
Most of the strategic and operative goals of the project have been achieved
The technological choices and strategic decisions for the optimal execution of the project’s activities and research have been always presented and agreed to the SMEs within the Consortium; their involvement in the development of the project work has been precious and always encouraged.Work progress and main technological achievements
WP1 – Concept Development

The WP1 was mainly dedicated to the identification of the main application scenario for the SMART-HIP project, to research and analyze the markets of interest together with the regulatory field as well as to identify requirements and specifications for all the components, which represent the basis for the product technical specifications.
The analysis of the application scenario identified, Hip Endoprosthesis, shows that, in the total hip replacement framework, there is an urgent need of the determination of implant loosening in detection method but, currently, the diagnostics used are not specific for the problem. The result is that the need remains currently unmet.
All SMEs, led by the WP leader MERETE, have conducted workplace studies and survey in the target application market supported also by the RTDs partners. A market study was carried out comprising the occurrence of inserted hip stems and acetabular cups in primary surgery. The market research includes data from the European framework and, then, the most interesting market segments have been investigated: Germany and Sweden.
The analysis of the state of art of the most common techniques currently in use for the implant loosening detection shows that there is a lack of specificity and, when the specificity is high enough, the technique is far to be a standard in the clinical practice due to the high costs associated.
The SMARTHIP device is able to raise the sensitivity and specificity of loosening diagnostics to a higher level, drastically reducing the occurrence of false positive or negative diagnosis permitting a high cost-saving rate.
Focusing on the use application scenario identified, a Requirements/Specification matrix has been implemented in order to map all the possible system configurations that can lead to a successful concept development. Therefore, during the three months of WP1 a preliminary concept of the system architecture of the Smart-hip project has been depicted, also in accordance with the involved European regulations regarding medical devices as well as Electromagnetic Compatibility (EMC). The system is mainly composed of three individual parts: Excitation Coil, Oscillator Unit and Detection Unit.
In addition, the integration hypothesis of each of the parts mentioned above have presented within this WP. It was a preliminary idea mainly regarding the possibility of merging together the Excitation Coil and the Detection Unit, which will be further developed during the second period of the project within the WP4 – System integration, working on optimization and taking into account the final user’s needs.
Finally, the Ethical Manual of the SMART-HIP project has been implemented, in which all the provisions necessary to duly collect, process, store and dispose of the anonymized samples that will be used in the project are described. It is to be noted that SMART-HIP will not carry out any trial on living humans or animals and the objective is to determine the accuracy of the new methodology, with absolutely no interactions with living humans or animals.
APPLICATION SCENARIO AND MARKET SEARCH
The first task of the project was managed by the Task leader (MERETE) with the proactive commitment of all partners. The consortium started the project with a workshop (held during the Kick off meeting) in which the main topics were the definition of the application scenario and the market segmentation, in order to address the project in specific markets segments and towards specific end users and stakeholders. The Kick off meeting was also the occasion for the RTDPs to start collecting the requirements of the SMEs and of the end users (thanks to the presence of expert people from MERETE which work in the field).
The specificity of the innovation permits the SMEs to focus the attention only on the application scenario of the hip prosthesis. As first step, an overview of the state of the art of the scenario has been reported. Today around 300 different total hip stems are available on the market worldwide. They are explored to resist high loads during different activities. According to the constitution of the patient, THR have to withstand up to two million load changes per year with forces of up to 870 % (during stumbling) of the body weight. Arising from this, high requirements are placed to the endoprosthetic implants. At the present time modular-built total hip replacement systems are offered almost exclusively. A modular THR system consists of an acetabular cup, a cup inlay, a femoral head and a hip stem. Each of the components can be chosen in different sizes and materials. This offers a more easily intraoperative adaptation of the hip endoprostheses components to the biomechanical circumstances. Moreover, they enable to choose between several tribological combinations.
Total hip arthroplasty became a routine procedure in orthopaedics. Independently of the type of the endoprostheses, the new ones reach survival rates of around 95 % after ten years, 84 % after twelve years, around 66 % after fifteen years and about 50 % survive for more than 18 years. The factors influencing the life expectancy are very considerable. They depend on demographic (age, sex), biomechanical (body mass index, acetabular inclination) and morbidity factors (cemented, cementless) as well as on the surgeon´s experiences.
Key factors influencing the life expectancy of endoprostheses are unknown to a wide extent. If the end of the life cycle is reached, a revision surgery is necessary. Studies have shown that in 6 % of the surgeries revisions are necessary after five years and around 12 % after ten years. An analysis of the revision surgeries after total joint arthroplasty was performed by Sadoghietal using the worldwide arthroplasty register and summarized the most common causes for revisions. They found out that the most common causes for revisions were aseptic loosening (55.2 %), dislocation (11.8 %), septic loosening (7.5 %) and periprosthetic fractures (6%).
Where pain is experienced following a total hip replacement (THR), there needs to be clarification as to whether the cause is due to an infected or mechanically loose THR. The optimal management in case of hip pain is an often discussed controversy. Currently, several diagnostic methods are used to identify the loosening status of the THR and to establish a basis for revision management. In general, diagnostic methods can be divided into imaging and implant integrated sensors. In general, at this stage there is no in-vivo system on the market, which allows the precise diagnosis of endoprosthetic loosening so far.
MARKET STUDIES
A market study was carried out comprising the occurrence of inserted hip stems and acetabular cups in primary surgery. The market research includes data from the European framework and, then, the most interesting market segments have been investigated: Sweden and Germany.
The growing volume of hip replacement is contributing to health expenditure growth since these are expensive interventions. In 2009, the estimated price of a hip replacement on average across European countries was about 8 000€. The SMEs will address the market segments in which they have already commercial business in place. However, with the purpose to expand the market share and access the others most profitable EU countries, the number of prosthesis sold have been estimated.
Internal sales statistic from the SME´s has also included. Referring an extensive market analysis, the general system has to be divided into two parts: the instrumented hip stem and the excitation-/evaluation unit. Thus, data was collected regarding devices for loosening diagnostics. A deep analysis is carried out in D1.1 comprising the analysis of the main barriers to entry and the main competitors.
REGULATORY ISSUES
A deep analysis has been reported in order to obtain the EU certification for the medical devices. The EU directive 93/42/EWG is the most important regular directive instrument for safety confirmation and medical technological performance of medical devices in the European Economic Area. The EU directive 93/42/EWG is the basic directive for Medical Devices Act in Germany. However, the Swiss framework has been taken into consideration, since Switzerland has taken over the European Union (EU) system of compliance assessment and certification, based on bilateral agreements. Switzerland has concluded treaties with EU Member States, EFTA States and Turkey with regard to the mutual recognition of conformity assessments for medical devices (bilateral agreements or mutual recognition agreements - MRAs). The basis for these treaties consists of the European CE mark and the application of the European guidelines on medical devices. The treaties simplify the mandatory reporting requirements for placing devices on the market, and permit direct distribution from Switzerland to all EU and EFTA Member States and Turkey, without needing to have an authorized representative with registered offices in the said countries.
USER REQUIREMENTS AND SPECIFICATIONS
According to the requirements of the SMEs partners, the device should be adjusted and fitted to the hip stems with a threaded bore hole for punching the hip stem into the femur. These bore holes are available in the cementless Z-Stem and the IntraBlock Twin Stem hip stems from Merete. According to the actual sales statistic, the Z-Stem would be our preferred hip stem for application and development of the diagnostic device. If the device could fit into a fixation screw, the MultiCup Locking PressFit cup would be the choice for acetabular cups.
In order to test the device on a cementless stem, the Z-Stem from Merete will be coated by Alhenia. The coating applied (Henia-Ha) consists of a hydroxyapatite layer and has the following characteristics:
1.Coating thickness: 100 µm -150 µm
2.Coating porosity: Max 10%
3.Coating roughness Rz: 50 µm – 100 µm
In addition of those requirements which represented the baseline for the RTDPs, the last part of the D1.1 named “technical framework”, is dedicated to the detailed list of the user requirements and specifications. The collection, analysis and coding of the requirements and specifications has been mane following the “IEEE Guide for Developing System Requirements Specifications”.
CONCEPT DEVELOPMENT
Here follows a short overview of the main outcomes for the different functional blocks of the system:
1.Excitation coil: the activities start off with Coil design, initially conceived with ‘C’ shape changing in cylindrical shape and still in progress with several hypothesis. Also the different materials (ferrite, iron) are taken into consideration taking into account benefits and drawbacks of each shape as well as different operating parameters (such as frequency, magnetic flux density, leakage inductance, etc..). Whatever solution, the excitation unit will be enclosed by an external case suitable for clinical practice, therefore it should be disinfectable and the materials used must be heat resistant, as the coil temperatures might rise.
2.Oscillator unit: the technical choices from UMR have been used by Labor to properly design the Oscillator Unit. This will allow to combine the test phase with the development phase which will lead to the real product. The Oscillator unit is a passive element which is excited by the magnetic field of the excitation coil. Being in contact with human tissue, the unit will be made by certified materials to ensure biocompatibility and all its components must be sterilizable as well as the screw head should not limit the range of motion of the hip endoprosthesis and therefore risk an impingement. For this reason a range of motion analysis was performed creating models to investigate the maximum dimensions of the additional top part and its maximum abduction of the hip endoprostheses. This led to the decision to design the screw without head as countersink screw that has to be inserted in the existing thread in the Z- hip stem from Merete in a way to allow an excitation vertically to the femoral axis. The sound is emitted to the surrounding bone and soft tissue. The size of the ball, placed in a cylindrical channel inside the screw, is limited by the screw size. For the first prototype a ball of diameter 5 mm should release an adequate impingement. The material of the ball is magnetisable steel or ferrite in order to allow a displacement from its starting position; magnetisable material instead of permanent magnetic material is used to allow other medical treatments and application of imaging methods.
3.Detection unit and accelerometer: the unit is composed by a vibration sensor placed in proximity of the skin and connected to a PC, by means of a custom electronics. It has the aim to detect and measure the sound signal coming from the Oscillator Unit. The most important part of the detection unit is the vibration sensor (accelerometer) which is responsible for sound detection. The sound signal generated by the Oscillator Unit, is in fact a solicitation vibration. Thus an accelerometer tuned on the frequencies involved is a proper solution in order to detect the signal. The analysis of the revealed signal should be correlated to the implant loosening, as many previous studies have already focused. Two of the most common technologies involved in the science accelerometer have been selected as proper constructive technology: MEMS and IEPE.
4.Acquisition chain: the vibration sensor is the first element of the acquisition chain and it is connected to an analog front end, which is responsible for post-sensor signal conditioning and Analog to Digital conversion (ADC). Afterwards the digital information has to be sent to a digital unit, whose core is represented by a microcontroller which is responsible for ADC interface and Device to PC communication.
5.Integration hypothesis: preliminary idea has been proposed, mainly regarding the possibility of merging together the Excitation Coil and the Detection Unit.
WP2 - Hardware Development
The design activities planned under this WP have been carried out taking into account the analysis developed in the submitted documentation, the D1.1 - Market study and Requirements Specification matrix and 1.2 - Product Concepts in particular. In the mentioned deliverables, the main system features have been deepened, drawing a first draft of architecture essential for starting the design activities.
The main blocks of the system are:
1.Oscillator Unit
2.Excitation Coil
3.Detection Unit
4.Personal Computer
Blocks 1 and 2 are responsible for sound generation, essential for the loosening detection. The Excitation Coil enables/disables the movement of a metal ball enclosed in a screw installed in the upper part of the prosthesis. That is the Oscillator Unit, which generates a sound signal correlated to the loosening.
Blocks 3 and 4 are the main parts of the Diagnostic Unit, wherein the sound is detected, computed and shown to the final user by means of a HMI (Human Machine Interface).
The architecture already presented in D1.2 has been substantially confirmed and both UMR and LABOR have worked keeping in mind those main ideas.
The unit was designed for the use in primary and revision hip stems. Most hip stems have a threaded borehole located proximally which is initially designed for attaching a removal tool in case of revision. These boreholes already included in implant design exhibit a simple opportunity to attach the excitation unit. The Excitation Unit is seamlessly integrated in a screw, consisting of a magnetic or magnetisable ball which is fixed on a flat leaf spring. An extra-corporeal coil, placed outside the patient’s body is used during the clinical test to excite the metal ball enclosed in the Oscillation Unit, which impinges inside and causes the implant to vibrate in its eigenfrequency.
Functional tests were carried out with an external coil voltage signal of +/- 7.5V 15Vpp. This signal was amplified by a power amplifier A1110-5 with a gain of 10. The distance between coil and oscillator unit was about 10cm. For the excitation a rectangular periodic signal was used, also an excitation by a sinusoidal signal was evaluated in additional tests. The experiments with the systems at a frequency of 5Hz have shown that the magnetic force is very weak over a distance of 10cm and more.
A second layout was drafted including the magnetisable ball with a fixed leaf spring in order to generate an oscillation which is directed linearly without a twist of the ball. With the use of this leaf spring, the direction of the magnetic field of the ball cannot be altered.
Pros:
Ball is led -> no free motion of ball
Cons:
Higher force needed to excite ball
Design is more fragile, more difficult to manufacture
Components:
Screw body and cover
Screw consists of 2 parts, body & cover
Material: Ti6Al4V
Weight: 1,3 g (body)
Weight 0.4 g (cover)
Magnetizable Ball
Neodymium NdFeB,
Surface coating NiCuCr
Weight: 0.5 g
Remanent flux density: 1.3T
Leaf Spring
Material: CuZn37 – CW508L (MS63) R480 (F55)
Thickness: 0.1mm
Weight: <0,01 g
With regard to our tests, the screw with leaf spring and magnetisable ball seems to be most promising, in spite of challenges. All components except the ball have been iteratively improved regarding easier excitation of the ball:
Socket in the cover that holds the spring was optimized
The length of the Spring was increased
The spring was designed to have a notch to decrease its stiffness
The inside of the screw body was optimized to maximize the possible length of the spring
EXCITATION COIL
The Excitation Coil is placed outside the patient’s body during the clinical examination and is used to excite the Oscillator Unit. Further drawings of the excitation coil were included in the related Deliverable, D2.1.
Subsequently, a pair of air coils was investigated. As our tests proved, this pair of air coils is the preferred design of all examined coil layouts. This geometry provides a magnetic force that is high enough to excite the magnetic ball from a longer distance. This was consolidated by COMSOL Multiphysics 5.0 FEA Simulation and experimental tests.
The finite element analysis (FEA) describes the electromagnetic coil and calculates the force between this coil and the neodymium ball. The system was simulated using the software COMSOL Multiphysics 5.0. As already mentioned, first preliminary functional tests could be performed and measurements could be recorded.
SIMULATIONS

The finite element analysis (FEA) describes the electromagnetic coil and calculates the force between this coil and the neodymium ball. The system was simulated using the software COMSOL Multiphysics 5. The following parameters were used for the simulations:
Ball
Material: neodymium
Mass 0.5g
Magnetic flux density 1.3T
Coil
Coil 1-4 (with values electric current, voltage, windings and geometrical parameters from the table 3.2 Electrical and magnetic specifications)
Material: copper
The material used for the coil cores is Iron Powder Vetroferrit (COMSOL 5.0) with additions of the materials MF 198. The original datasheet from TRIDELTA is reported in Annex II.
Environment
Material: air and water
Distance x: 40 – 110mm

FUNCTIONAL TESTS
For preliminary functional test, the following experimental setup was designed and constructed according to DIN ISO 2768-m. The test system consists of several components:
1: Total hip replacement (THR)
2: Embedded in an artificial bone cylinder, Sawbone 20PCF (Sawbones, Vashon, USA).
The z-stem is integrated into a Sawbone cylinder and the cup in a Sawbone block. Additional tests were executed with an artificial bone cylinder, Sawbone 10PCF.
3: Water as a tissue replacement
4: Latex cover as skin tissue
5: Variable weight load on top
In addition, this experimental setup device is also equipped with several mechanisms for a decoupling from the environment. These consist of several rubber elements and a marble plate.
DETECTION UNIT DESIGN
The Oscillator Unit (activated by the Excitation Coil) produces a sound vibration which contains the useful information leading to the loosening status. The Detection Unit is, therefore, responsible for the detection of such sound and its consequent digitalization.
The final architecture (discussed below) is the result of the design activities carried out by LABOR, during the two working years, whose results have been shared with the whole Consortium and discussed in several meetings and technical exchanges. The results have been eventually described in the related technical deliverables, as well as in the two Periodic Reports, already submitted.
The main parts of the Detection Unit are roughly interconnected:
• Sound detection and signal conditioning: the sound signal is detected by a proper sensor (IEPE accelerometer) and, by means of proper signal converters and conditioner, is then prepared for the A/D conversion;
• Analog to Digital conversion: the signal coming from the signal conditioning stage is carried to the A/D converters, responsible for signal digitalization;
• Data packing and sending: the raw data packet supplied by the A/D stage is prepared and sent to the PC by means of an ad hoc protocol;
• Management of the Excitation Coil: the excitation of the Oscillator Unit – provided by the Excitation Coil – has to be synchronized with the activation of the acquisition stage, in order to fulfill the correct timing of the whole measurement.

Such architecture allows to focus computational activities and signal post-elaboration on the Personal Computer, as the output is a raw data stream representing the sound information coming from the Oscillator Unit, in turn already enabled/disabled by the Excitation Coil, which is controlled by the Detection Unit itself, and powered by means of a proper circuit, the Coil Driver (see below).
As far as sound detection is concerned, it was agreed that an IEPE accelerometer – specifically the Metra KS95B100 – was the best choice in order to fulfill our requirements, e.g. wide bandwidth, low noise, reliability (see D1.2 for further details). Moreover, the mentioned sensor had been successfully used in the experimental setups which led to the present Smart-Hip project, so we thought reusing the cumulated skills and experience was the best approach in order to optimize the design activities.

At first, it was envisaged a first prototypal setting, wherein, due to reasons related to quality of transmission, the whole circuit would have been split in two sub-blocks, namely:

1.The Sensor block, composed of the first three stages of the Block diagram reproduced above:
Accelerometer
IEPE converter and Signal Conditioning
Analog to Digital Conversion
2.The Control block, composed of the components responsible for computing and communication:
Microcontroller
USB Data interface
Therefore, in order to protect the acquisition stage from external noise, possibly routed by long cables (i.e. sensor’s wires), the Sensor block was planned to be positioned as close as possible to the accelerometer, on the leg. For that reason, the accelerometer itself, the signal conditioning stage as well as the A/D converter have been thought to be placed within the same enclosure, to be digitally connected to the next Control block by means of a proper shielded cable, suitable for the present kind of application.
The Control block would have been placed just outside the area where the Excitation Coil as well as the Sensor block operate. For instance, the Control block could be placed on a desk, close to the Personal computer.
Such solution was abandoned over time because the preliminary tests showed that it was not possible to reliably drive the sound digital signal by means of a very fast digital interface (i.e. SPI) over a quite long cable which would have connected the A/D stage (over the body) with the microcontroller (on the Control Block, on the desk). That is the reason why we decided to keep the whole Detection Unit within a single board.
The accelerometers, therefore, are now the only components applied to the leg and their own cables carry the IEPE signal to the signal conditioning input, on the main board. Measurements of signal integrity led over the mentioned IEPE cable demonstrated that the sound signal is clear enough and noise-free, so confirmed the correctness of our choice.
FIRST EXPERIMENTAL PROTOTYPE AND TESTS
The first experimental prototype was realized on two stripboards, equipped with the blocks already mentioned. The first stripboard was devoted to testing over the IEPE conditioner/amplifiers; a Discovery STM32 board (for firmware development and evaluation) overlaps the second stripboard which contains the ADC section as well.
Such components were interconnected by means of special wire, usually employed in the electronics prototyping activities. The stripboard was, in fact, an experimental “forefather” of the next PCB (Printed Circuit Board) realized in the final part of the WP2 and developed after stripboards had evaluated the adopted solutions.
The stripboards have been built taking into account modularity and easiness of development. For example, a ground grid has been realized in order to produce a kind of ground plane and at the same time to bring the GND nodes along the whole circuit, where needed. The same criteria have been adopted for the several power supplies, brought from the source to the desired spots.
The Discovery board includes the microcontroller Unit. It has been connected to the data interface and to the analog front-end as well. Such board has been used for the preliminary firmware development. Besides, it relies on “Schmitt trigger” oscillators, designed in order to simulate the input signal downstream the IEPE circuit.
FINAL PROTOTYPE AND TESTS
After first preliminary tests over stripboards were completed, LABOR was ready to start designing the final PCB, developed taking into account the experience cumulated during the preliminary prototyping phase. Such design activity led to the final release of the 2-layer PCB, which included, on a single quite small card, the electronic blocks already mentioned.
COIL DRIVER
Although not mentioned in the DoW, a Coil driver was strongly needed in order to replace the bench power supply, which could not be integrated with the Detection Unit. Without such Driver, the integration activities would have been very difficult and the final system not completely exploitable as a custom product.
Therefore, in order to fix that issue, LABOR gathered the main electrical specifications (waveform, amplitude, power etc.) and designed an integrated unit, to be used along with the Detection Unit and directly controlled by the latter.
The Driver is a MOSFET-based class B amplifier, with one op-amp able to accept high voltages, up to 140 V. The input signal is provided by a DAC signal line, previously prepared on the microcontroller of the Detection Unit. It is possible to control the amplitude as well as the waveform (i.e. square or sine wave), directly by modifying the firmware code. The design activity led to the construction of a complete unit, composed of a custom PCB, enclosed within a plastic case, the same as the one already used for the Detection Unit.
PERSONAL COMPUTER
The personal computer or end user HMI for the project has to be chosen according to several requirements.
1.The first requirement is the good communication via hardware port according to the real and definitive hardware board design. This feature is important to ensure good reliability of the signal received from PC, so to have accurate data elaboration and the best response, close to reality as good as possible.
2.The second requirement is the good usability by the end user point of view i.e. the Doctor. This implies to choose between touch screen or not, right size and resolution and the chances for housing in the rest of the Diagnostic unit design.
Regarding the first requirement, several types are available on the market: in fact a reliable communication port can be:
1.Ethernet: standard and industrial pc port, distributed options are available
2.RS485: industrial standard port, support long connection and many automation devices
3.RS232: standard port with industrial purposes in many cases, strong communities support
4.USB: standard port with high throughput in the latest version, useful for non real time app
5.MODBUS: industrial port largely used in HMI where devices don’t have standard screen
MECHANICAL DESIGN
The mechanical design required an important effort, in order to provide a suitable prototyping enclosure for the Detection Unit. It was chosen a plastic box, where the PCB and the power supply stage were enclosed in.
Three BNC connectors (one per accelerometer) have been placed on the front panel and they connect the sensors directly to the board. Beside the PCB a proper power supply stage has been placed. The rear panel features a second small PCB, containing the USB connector as well as the Coil control: by means of a switch – specifically by means of a pedal switch – it is possible to activate/deactivate the signal output directed to the Excitation Coil, so as to synchronize the excitation of the Oscillator Unit with the sound acquisition.
Alongside this, several overall solutions were examined, taking into account the main features that the medical environment usually requires. Among several hypothesis, it was assessed the possibility to place the first prototype of Excitation Coil on a special arm, free to be moved over the patient’s leg to seek the best excitation spot. Such hypothesis was formulated with the possibility to use the mentioned two blocks (Sensor and Control), so the Sensor Block was expected to be placed on an elastic belt, whilst the Control Block was planned to be positioned outside, possibly placed on a desk, in proximity of the PC, where the Diagnostic Software would run
Even abandoning the two-block splitting, the mentioned hypothesis is still valid with the single unit (see above) placed on the desk, while the latest models of Excitation Coil should be adapted according to such solution.
CONCLUSIONS
At the moment we write, the whole system can be considered at an advanced stage of development. The mock-up is available for test sessions like the ones carried out by UMR in Rostock under WP5, during which the whole system (Oscillator Unit, Excitation Coil, Diagnostic Unit) were tested. The interaction between the parts already developed and built, represents an important result studied and deepened under WP5. Summarizing:
The Oscillator Unit has been produced in two different layouts:
1.The first one is based on a loose and freely movable magnetisable ball inside the screw as described by Ruther et al. 2014 (UMR).
2.In the second layout the magnetisable ball is fixed with a leaf spring in order to generate an oscillation which is directed linearly without a twist of the ball.
The screw with leaf spring and magnetisable ball seems to be most promising, despite challenges. All components except the ball have been iteratively improved regarding easier excitation of the ball.
The Excitation Coil design focused on a pair of air coils. This is the preferred solution among all coil layouts examined so far. This geometry provides a magnetic force that is high enough to excite the magnetic ball from a long distance. The related power supply stage was designed and prototyped, according to the coil’s electrical characteristics.
The blocks belonging to the Detection Unit have been prototyped. Therefore, LABOR developed a complete device, turning the first stripboard prototype into a final complete prototype, composed of:
1.tailor-made PCB (Printed Circuit Board)
2.embedded power supply stage
3.custom plastic case
The PCB contains all the components previously tested by means of the prototypal strip-board, namely:
1.IEPE converter
2.A/D stage and signal conditioning
3.Microcontroller and fast data interface
The Detection Unit is equipped with 3 IEPE inputs (front panel), in order to support UMR with the possibility of using three accelerometers. Besides, the Unit gives the opportunity to control an external Coil unit, by means of a pedal.
Finally, the mechanical aspects have been deepened so a proper plastic case, equipped with LEDs, connectors and other connections, has been provided.
Design activities:
1.Analog stage design: IEPE accelerometer and generator, conditioning stage, ADC block.
2.Digital stage design: microcontroller unit (ARM architecture), real time data streaming stage (FTDI FIFO)
3.Coil Driver design: gathering of requirements, in collaboration with UMR: electronic design;
4.Schematic design and testing on stripboards.
5.PCB design.
The Detection Unit, as well as the Coil Driver, have been designed taking into account the system specifications so the design activities have been completed on-time and with best results in terms of reliability and product-oriented architecture. The tests in laboratory carried out on the Detection Unit as well as on the Coil Driver, have produced a positive outcome: in fact, they completely fulfilled the system specifications gathered and discussed during WP1.WP3 - Software Development
DIAGNOSTIC SOFTWARE
The activities related to the implementation of the diagnostic software, carried out under WP3, started from the following points:
1.Control of the ADC (analog-to-digital converter) and DAC (digital-to-analog converter) card NI USB-6216 (National Instruments, Austin, USA)
2.User input
3.Visualization
4.Recording of the digital measured values
5.Analysis of the digital measured values
as well as the storage of the data. The whole implementation was realized with LabVIEW 2013 (National Instruments, Austin, USA). As a suggestion for this process, first experiences from the WP2 Hardware Development, where several measurements have already been done, were included.
The signal processing algorithm interfaces with the HMI (Human Machine Interface) from work package WP3 Software Development. On the one hand different settings for the measurement are provided to the signal processing algorithm, e.g. the duration and power of the coil excitation. On the other hand, the processing algorithm sends different information like data analysis results to the HMI.
There are two main tasks that the diagnostic software has to accomplish:
1.Data acquisition and analysis
2.Generation of the signal for the excitation coil
The data acquisition has to collect the sensor data and the excitation signal as well to ensure synchronicity. The measured data is then processed. Several filters can be implemented, to increase the signal to noise ratio and to decrease the influence of parasitic errors like the 50Hz power line frequency.
Furthermore, a windowing of the time signal can be used to omit transient oscillations at the beginning of an excitation. The filtered time signal is then transformed into the frequency domain using the FFT (fast Fourier transform). The analysis of the signals is then carried out. Therefore the frequency spectrum and the time signal have to be evaluated. The analysis is based on data that has been collected during the field tests.
As a starting point a program was developed in LabVIEW 2013. This integrated development environment (IDE) provides in connection with the LabVIEW based data acquisition device NI USB-6216 a simple possibility to collect and store analog data. In this way, an activation of the excitation coil via the software is also possible.
Immediately after the measurement, the measured values are shown in a graph for a first evaluation. The frequency spectrum is also graphically represented, after the FFT operation has been performed. In the display of the measured values all three channels, the excitation signal and the trigger signal are shown. The FFT function display shows only the three channels and the excitation signal. It is possible to manually zoom in a particular area or to hide individual channels for a better view.
To analyze the digital measurements, two variants are provided. The first variant uses services of the LabVIEW module Spectral Measurements and the second variant uses a fast Fourier transform (FFT) algorithm with additional settings options.
The construction of a data file is divided into three sheets:
1.Sheet 1: In the first sheet, all relevant parameters are stored. This includes following attributes:
Name of the measurement
Parameter of all three channels (sensor parameters)
Parameter of the DAQ
Sample frequency, number of measured value and time interval
data type
Sheet 2: In the second sheet all measurement values and a time code are stored.
Time [s]
Sensor channel 1-3 [V]
Excitation unit signal [V]
Trigger signal [V]
Sheet 3: In the last sheet the first performed FFT, as mentioned in chapter 2.5.1 is stored.
Frequency [Hz]
Value (sensor channel 1-3) [dB]
To get further information that can be used for analyzing the measured data, computational analyses are done on frequency and attenuation. To prepare the simulations several 3D-parts were modeled. The parts allow the generation of different areas of loosening between Z-Stem and artificial bone. A mantle surface was created that represents the contact surface between stem and artificial bone. After the generation of a defect it is merged with the cylinder again.
For a Modal analysis, the first 15 eigenmodes of the specimen were simulated.
This includes bending, torsion and compression/tension mode
The first 15 eigenmodes are in a range from ca. 45[Hz] to 3.5[kHz]
The software that has already been realized in LabVIEW provides a sufficient fundament for the test of the loosening analysis. As already mentioned, first preliminary functional tests could be performed and measurements could be recorded with the preliminary software version and modules.
In addition, all variants, modules and options should be integrated in a single software package. This allows a simple operation and handling of the work package WP5 Field Tests.
SMART HIP SOFTWARE
Diagnostic Software:
The diagnostic software is a key part of the SMART-HIP project. Its purpose is to generate an appropriate signal for the excitation coils, also to acquire the signals outcomes from the sensors. The sensor signals are analysed regarding attenuation properties and frequency spectrum.
The diagnostic software has to present to HMI the analysis results that are available to the user.
The hardware interface which allows software to connect with the sensors and the coils is the data acquisition device NI USB-6216 by National Instruments (National Instruments, Austin TX, and USA). It provides 16 bits resolution DAC and ADC. The maximum sampling frequency is 400 KS/s (aggregate) for all used analogical inputs and 250 KS/s (aggregate) for all used analogical outputs. The device provides 8 differential analogical inputs (BNC) and 2 analogical outputs (BNC). The maximum voltage is ±10V in input as well as in output.
The activities of the diagnostic software can be listed as follow:
1.Excitation signal generation for the coil (DAC)
2.Sensor signals acquisition (ADC)
3.Signals visualization (for standalone usage without HMI unit)
4.Storage and Recording measurements
5.Data analysis
The current UMR diagnostic software is written in LabVIEW. The software is divided into two separate program sections: the first LabVIEW routine, MeasurementSigGen.vi is used to generate the excitation signal and to perform the measurements. The acquired measurements data are stored on hard disk TDMS file type (National Instruments properties). A second LabVIEW program, DataAnalysis.vi can easily read and elaborate the data format produced with the same format.
The first routine, MeasurementSigGen.vi is directly involved with the NI USB-6216. One side the excitation signal is sent to the coil in output, one side the sensor signals are read in inputs.
Several functionalities are implemented:
1.Signal generation for Coils
2.Signal live preview
3.User inputs start control included
4.Outcomes sensor signal preview with frequency spectrum
5.Recording data options to separate files
As introduced in deliverable D3.1 the level of loosening of a total hip implant influences its own frequencies. The changes of the spectrum can be analyzed in two ways:
1.Detection of shifts of frequency peaks
2.Detection of a shift of the spectrum’s area centroid
The first method implements a peak detection algorithm. It detects the position of peaks in the spectrum. A first draft of this algorithm is currently implemented as a Matlab script and can easily be acquired ore translated in LabVIEW.
The second method shows the spectrum with linear amplitude. First the area between the curve and the frequency axis of the graph is computed, and then the centroid of this area is calculated. This algorithm is also currently available as a Matlab script and can be acquired or implemented in LabVIEW since National Instruments acquired the rights for Matlab and a simple script node can import MATLAB code exposing inputs and outputs.
In deliverable D3.1 a time domain analysis approach was presented: in this analysis an approximated envelope curve has to be fit into the fading time signal. The parameters of this curve have to be analyzed regarding the decay properties and the damping coefficient. Currently detailed measurement data has to be obtained from field tests to gain the ability to supply a statement on loosening states.
The HMI:
Since the SMART-HIP machine is implemented on a high-speed electronic board, the data acquired by the system needs to be elaborated and presented to the final user, doctor or technician. Moreover the right amount of data and the right modality has to be chosen to have a perfect balance between the details of acquired signals and the data results views.
Generally speaking, the hospital itself is not an easy environment for technologies for several reasons (legal, profile of the users, peculiarities of the technologies in use and the responsibilities that those are called to manage). Furthermore, physicians profiles are in the middle between a normal user and a technician, therefore the software interface (HMI) has to be able to give sufficient information about technical details of the analysis and clear presentation of the significant data and results. At the same time the HMI software must let the doctor the final decision for the diagnosis.
The core of the system is the electronic board. However, since the board cannot perform a post-analysis of the data collected, and this analysis has to be done on the HMI, a PC/laptop/tables with sufficient calculation power is needed for the system. Generally speaking, this has to be done by a PC based machine with standard Operative System and standard Input output peripherals.
Further, the format and the placement of the PC make the difference: in cases in which the computers are directly involved in instrumentation, the best practice is to choose wall mounted touch panel with brackets or structure with armed equipped touch panel.
Regarding the physical connection, the system requires two standard USB connections: one for the Command stream and one for the Data stream. The Command stream is not a critical communication port, since few commands are needed to start the proper device action (as ‘turn on’ and ‘turn off’ the coil, ‘enable data stream’ and so on).The data stream port is a critical port, because the amount of data stream is actually considerable. In fact, we have three accelerometer sensors with sampling rate at 140 KHz. This means 140.000 samples per second on 16bit signed integer i.e. 280kByte per second per three channels, therefore, 840kByte per second. For this reason, the difference between the two ports is huge and this aspect has to be taken into account while dimensioning them.
In the first release of the system, the PC hosting the HMI will be separated from the electronic parts, so the physical connection will be made manually before or after software starting (detailed in the next sections) by two micro-USB cables, one for the data streaming port, one for the command port, as described above. The rest of the system, belt, coils, sensors, will be responsive to this two connection channels.
The software starts with its main form layout, depicted in the screenshot reported below. As it can be observed, the user has several controls and indicators.
In the upper left side of the form, two indicators show basic information related the status of the connections and the available power.
In particular, the ‘‘connected’ indicator (first on the left) shows if the command port is successfully connected. The “quality” indicator lets the user know how much power is available. Actually, the quality indicator is also a way to know if the data stream port is also connected. In that case, in fact, the power level would be at its minimum. However, this is not the right and complete procedure to follow to check if the signal is powered enough.
Right after the “connected” and “quality” indicator, the “select patient” section is made up by two buttons. Those are for the patient loading: the first one (“new”) is for creating and recording a new patient. The second button (“from list”) is for retrieving a recorded patient from list. By clicking one of those two buttons, two different forms will open, as described hereinafter.
On the left side, three important buttons appear: the Start button, the Test button and the Save button. The first one is to perform the measurement, the second one is to test the signal power, after turning on the coil, while the “Save” button allows to save the measurement performed. These three buttons constitute the core of the SMART-HIP software functionalities.
The acoustic model
Simulations provide a good initial base for basic research and parameter search, which were used in the task “Signal Processing Algorithms”. The finite element method (FEM) software COMSOL Multiphysics 5.0 (COMSOL Inc., Burlington MA, USA) and Abaqus (Abaqus 6.12-2 Simulia Dassault Systèmes, France) was used for various simulation calculations. COMSOL provides a simple mean for the creation of a Total Hip Replacement (THR) model. These models, thanks to their constructive characteristics can be designed/modified arbitrarily deep and complex in 2D or 3D environment.
The simulations can be divided into three categories:
1.Modal analysis
2.Attenuation analysis
3.Analysis of sound propagation
The first one, a modal analysis was performed a modal analysis with different setups. The THR was simulated in a test environment similar to the experimental test setup, in addition the influence of patient weight was investigated.
Using the same model the second analysis has been calculated. The study was mapped from the frequency to the time domain and the vibration behavior of the THR was analyzed in a certain period. From the location dependency of a predefined point on the THR, the damping coefficient of the system was determined.
Analysis of sound propagation represents the third simulation. In this simulation a 2D cutting plane of the whole system was created and the parameters of the individual materials were allocated in the study.
The results of all simulations should be compared with the results those obtained by experimental measurements and the parameters in the simulation and the experimental set-up should be gradually adjusted. This gives us a better conformity of theoretical model and experimental setup.
In the modal analysis the first 15 eigenmodes are determined. This includes bending, torsions and compression/tension mode.
Subsequently, a time dependent analysis was realized to investigate the damping. The attenuation analysis simulation evaluates the first 1.5 sec of a forced vibration, contrary to the Modal analysis, in which a frequency depended study was used, to analyze the first eigenmodes.
The simulation of sound propagation was performed in Abaqus.
The parameters of each material must be determined more precisely to customize the simulation even more accurate compared to the experimental setup. The first results of the WP5 Field Tests need to be compared and validated with the calculation of the simulation. Both methods need to be adjusted to one another to improve the simulation and the experimental work. In an ideal case, the results of the simulation and the experimental values agree at the end of the adjustment process.
There are also future simulation plans to design different setups and to compare them. This applies for example to the position of the patient (e.g. standing, sitting or lying) or the orientation of the oscillation unit (ventral-dorsal or medial-lateral). It has to be compared, how the frequencies of the eigenmodes, the damping coefficient and the propagation of the sound change in the specimen.
Another step for an accurate detection of loosening in a THR is the adaptation of the geometry to real patterns. That means the homogeneous SAWBONE cylinder is replaced by a SAWBONE artificial bone with realistic anatomy of cancellous and cortical wall. Furthermore the defects areas should be designed according to realistic defect models (e.g. AAOS classification or Paprosky classification).
The Simulations that have already been realized in COMSOL and Abaqus provide a sufficient fundament for the test of the loosening analysis. First results show a likely behavior and show a similar trend as in other works such as Ruther, Schulze which were calculated in preparations for this project.
WP4 - System Integration
The T4.1 T4.2 and T4.3 were devoted to System integration, Preliminary tests and System debug. During this WP the three main components (Oscillator Unit, Excitation Coil, Diagnostic Unit) were integrated and made working together, testing the overall operation and addressing issues arisen in the meantime.
Both the Oscillator Unit and the Excitation Coil were prototyped and tested.
LABOR shipped both the Detection Unit and the Coil Driver to UMR, in order to include the two devices into the experimental chain installed in Rostock, alongside the existing bench test setup mainly composed of bench IEPE amplifiers and coil drivers, as well as National Instruments embedded electronics and software. The SMART-HIP Detection Unit and Coil Driver have also been validated in comparison with that bench equipment. Finally, the two SMART-HIP custom units were tested within the final test setup and some minor issues were isolated and recorded.
Therefore the prototype was ready to be tested. The Coil Driver was finalized and produced as soon as WP4 gave a positive outcome and stable specifications were available. Right afterward the Excitation Coils was provided with its own power supply module perfectly integrated in the system. While the power supply for the coils was under construction and no output channel was available at the moment, UMR software and hardware was used for signal generation for the purpose of preliminary tests.
Furthermore a connection to the measurement PC running the HMI was established. After some minor modifications to the HMI and adjustments of the hardware measurements could be conducted to test the interaction with the Oscillation Unit and the Excitation Coils.
At first the Diagnostic Unit is compared to the measurement hardware that was used in the test setup at UMR by now. UMR used the National Instruments data acquisition box NI USB-6216 BNC for measuring and signal generation. The device has a cumulative maximum sampling rate of 400 kS/s. There are five of the eight available analog input channels in use, three for the sensor signals, one for trigger signal and one for the excitation signal. Each input channel is adjusted to use a sampling rate of 80 kS/s. For the generation of the coils’ excitation signal one of the two available analog output channels is used. The resolution of the analog digital converters is 16 bits.
The HMI is a LabVIEW program that is designed for conducting measurements and for organizing a patient database.
The data communication to the PC is realized with a USB/UART connection. Therefore the Diagnostic Unit is equipped with the USB to UART chip FT2232H by FTDI. This chip offers two serial ports. The HMI uses both ports, one as a control channel and the other for retrieving measurement data from the connected sensors. During the preliminary tests all three sensors are used. As a preparation it is necessary, that the FTDI driver is provided for the data channel and that the driver NI VISA is installed for the use of the control channel.
In order to be able to compare the measurement results of the Diagnostic Unit to those of the UMR setup several test were performed. The sensors were mounted directly on aluminum pads which are glued to the artificial bone (Sawbone cylinder). This connection allows dry measurements and ensures an appropriate reproducibility.
The Sawbone cylinder is equipped with twenty aluminum pads. There are five pads located equidistant on every side of the cylinder. The sensor positions are named after the sides (medial – M, lateral – L, anterior – A, posterior – P) and the number beginning from the top. For the preliminary tests the three sensors were located at the positions A1, P1 and P3.
For the tests an excitation signal with square waveform and a frequency of 98.5 Hz was used. At first, measurements were done with the UMR acquisition hardware. Then the tests were repeated with the integrated Detection Unit. The sensor position and the overall test setup were not altered between the measurements.
PREPARATION AND RESULTS
Several preparations had to be done in order to compare the measurement results of the Diagnostic Unit to those of UMR. For the evaluation of the measurement data the data analysis software of UMR is used. As the HMI produces measurement files (TDMS format) that have another structure than the ones produced by the UMR measurements a conversion is needed. For this purpose a LabVIEW program was created.
Several measurements were conducted, both with Diagnostic Unit and UMR setup. The results of the preliminary tests validate that the Diagnostic Unit works well together with the test setup at UMR. The good accordance of the measurements shows that it could replace the UMR measurement hardware. This is a first step towards the goal of having a fully integrated measurement system without the need of additional hardware like a charge amplifier for the sensors.
SYSTEM DEBUG
During the system integration, some adjustments had to be done to both, the HMI software and the Diagnostic Unit hardware. Before the HMI was ready to run, some issues had to be fixed.
1.The first problem found during debugging was connected to the date format. There was an inconsistency, sometimes the format DD/MM/YYYY was used and in some cases the format DD.MM.YYYY. A conversion that replaces dots with slashes was implemented into the SubVI “Patient.vi”. This solved the problem.
2.Another problem found during debugging was that the list of patients could not be loaded. The list was always empty, even if there were already patient files in the folder. The solution was to alter the filtering for TDMS files in the SubVI “Patientlist2.vi”.
3.Furthermore slight adjustments to the Diagnostic Unit had to be done. On the one hand the signal amplitudes of the sensors attached to the artificial bone in the test setup were too high. To avoid clipping of the signal the amplifier gain had to be reduced. This was done separately for each of the three sensor channels by adjusting the corresponding trimmer on the PCB. On the other hand there was an offset on the signals that also could be corrected by adjusting the trimmers.
4.Another problem found is short disturbance on the signal that occurs randomly from time to time. The error is visible on all sensor channels simultaneously whether or not a sensor is connected. Debugging was planned to find the reasons for this issue.
CONCLUSIVE REMARKS
The Diagnostic Unit was developed and built by LABOR. It was sent to UMR to integrate it into the test setup including the Excitation Coils and the Oscillation Unit. The system integration was done at UMR. At first, and while the Diagnostic Unit’s power supply was under construction, the signal generation was performed with UMR’s equipment. As soon as the power supply was completed, the Excitation Coils was then activated by our custom electronics, developed according to the power supply currently used at UMR.
After a debugging period and some adjustments the system was made operative and measurements could be conducted. To compare measurements of the new system to those of the former UMR measurement hardware a conversion tool was also developed. The comparison of the recorded data shows a very good similarity.
Despite some small issues that still have to be removed the Detection Unit proved to work well together with the test setup at UMR. As the HMI’s purpose is to be used from end users (e.g. doctors...) it is not adapted for doing fundamental field test. For this purpose the UMR software is better suited and will be further used for the ongoing debugging and fundamental tests. The partners assessed that each planned improvement or progress can be easily implemented in the HMI later on.
WP5 - Field Tests
DESCRIPTION OF TESTING METHODOLOGY
For the conduction of the experiments a mechanical test stand was built. It allows the fixation of femoral bone and is equipped with decoupling against disturbing vibrations from the environment. The stand mimics the patient’s leg including a total hip replacement with a hip stem and an acetabular component.
As a first series of tests with the SMART-HIP diagnostic system, we used specific cylinders of closed cell polyurethane (PU) foam (Sawbones® Europe AB, Malmö, Sweden) with a density of 10 pcf (0.16 g/cm3) as a substitute for the femoral bone. The used cylinders had a length of 30 cm and a diameter of 5 cm.
With each of the artificial and human femur specimens, three different loosening states were examined consecutively. The first one, called L0, is the initial pressfit-state obtained after implantation. A set of measurements were conducted before the subsequent loosening state was prepared. Thereby, L1 has a depth of 3 cm and L2 has a depth of 6 cm.
For evaluation of the sound signal, the Z-Stem manufactured by partner MERETE was used in combination with a metal ball and an acetabular component. The Oscillator Unit was screwed into the threaded borehole of the stem. Sound excitation was achieved by placing an electromagnetic coil next to the stem.
The investigations using artificial bone cylinders were conducted firstly under dry and secondly under wet conditions. Opposed to dry condition testing, a water bath as substitute for muscle and fat tissue was used in wet testing. Thereby, a latex tube functions as a boundary for the water bath and also as skin replacement. Within the dry conditions, the sensors are placed directly at the bone, while within the wet test environment, the sensors are applied directly to the latex tube as planned for real application on the patient’s skin. The sensors were attached either with a double side duck tape or highly elastic rubber sleeve. For measurements in wet conditions, the single electromagnetic coil from dry testing was replaced by a pair of air coils with each 1200 windings.
The duration of each measurement is five seconds. Only during this time the excitation of the magnetic field is activated and the magnetic ball is oscillating generating vibration of the implant. The LabVIEW program is able to record the sensor signals and to save them to a measurement file. The used sample rate of the analog digital converter is 80,000 samples per second for each of the three sensor channels. A second LabVIEW program is used to pre-analyze the recorded measurement data. It allows to select a specific time portion of the recorded signal and to do a Fast Fourier Transform (FFT) for getting the frequency spectrum of the signal. For each impulse a sampling duration of 10 ms is chosen because the signal is faded to zero after that time and the next impulse follows after that because of the periodical excitation. With a sampling rate of 80 kHz the resulting number of samples per impulse response is 800. After normalization and transformation to frequency domain a number of 401 frequency bins results with a maximum frequency of 40 kHz. Because of the expected interesting area of the spectrum being not higher than 20 kHz we decided to do a resampling by the factor of ½, resulting in a number of 201 frequency bins with a maximum frequency of 20 kHz. At first a principal component analysis (PCA) is used to reduce the dimensionality of the data set from 201 dimensions to 3 principal components. That means that each measured impulse response is now represented by 3 values instead of the 201 frequency bins. The so collected and transformed data set is used as a training basis for the machine learning algorithm support vector machine (SVM).
RESULTS FROM DRY TESTING USING BIOMECHANICAL BONE SUBSTITUTE CYLINDERS
Within the dry measurements, we recorded a total of 121 datasets, containing 46 records with a press-fit state (L0), 15 measurements with a small defect (L1) and 60 measurements with a larger defect (L2). The resulting datasets all have a duration of 5 s including 4 impacts. The time signal of the impact signal portions is transferred to the frequency domain with a result of 201 frequency bins, which are reduced to 3 principal components. The PCA is able to cover 96% of the variance in the data in 3 Principal Components. The data are randomly split into two subsets, a training subset and a test subset. The PCA and SVM (with a linear kernel) are applied on the training subset. This results in hyperplanes that are able to divide all three classes linearly.
The output of the classification report shows that 100% of measurements of the test subset are predicted correctly by the classification done by the SVM.
RESULTS FROM WET TESTING USING BIOMECHANICAL BONE SUBSTITUTE CYLINDERS
Additional measurements using Sawbones models were performed in a wet condition. The water is intended to simulate the soft tissue, fat and blood. A latex cylinder containing the water represents the human skin.
A total of 1644 measurements were recorded, 644 for press fit L0, 480 for small loosening L1 and 520 for larger loosening L2. All measurements were processed as measurements in dry condition whereby the sensor mounting was different. All three accelerometer sensors were attached to the latex cylinder with the use of a rubber band or Velcro tape. In the whole set of measurements, a recognition and classification of all three states of loosening L0, L1 and L2 could be performed 100% correctly.
RESULTS FROM WET TESTING USING HUMAN FEMUR SPECIMEN
The in-vitro tests with human femur specimens were conducted according to the standard operating procedure (SOP) presented in Annex I of deliverable 5.2 (All human femur tests were run under wet conditions as described in Deliverable 5.2. A total of four human femora were obtained from the Institute of Anatomy, University Medicine Rostock and used for testing of the Smarthip system.
In the following, the results of each human femur specimen are presented in terms of classification report as score plots. In addition to the procedure under wet conditions using artificial bone, the sensors were applied using 3D printed ABS sensor mountings. These were attached to the latex tube with 3M double-sided tape. Thereby the same position and contact pressure are ensured for all measurements.
Pull-out tests were performed on the human femur specimen extracting the endoprosthetic stem from the bone bed using a universal testing machine. Significantly lower pull-out forces were measured with larger level of loosening.
In addition the extraction test, the total hip replacements were compared in x-ray images.
Before the preparation of the human femur, the implant suitability with regard to the geometry of the human femur was examined.
The UMR software detection part and Signal generation were replaced by the detection unit and the HMI components, produced by LABOR. Recorded data are in both versions almost identical with very small differences.
Comparison of the parameters of both variations, indicate that the LABOR-HMI version is equal or better in all variations.
In preliminary tests in combination with the Detection Unit, both systems were compared with each other.
Recorded data are in both versions almost identical with very small differences. The new variant, LABOR – Detection Unit is even more precise. The calibration of the accelerometer sensors is ensured with the certification of PCB Piezotronics.
ECONOMIC ASSESSMENT OF THE SMART HIP TECHNOLOGY
The economic assessment of the SMART HIP Technology was detailed in Deliverable D5.3. This shows the activities performed to assess the economic feasibility of the solution, providing information on:
1.The SMART-HIP Diagnostic Device (meaning Extracorporeal coils, Measurement device and Coil driver),
2.The SMART-HIP Software
3.The SMART-HIP Intelligent Prosthesis (meaning Oscillator Unit and Hip Stem)
Part of the task was analysing the manufacturing process of each component of the SMART-HIP System, based on the exploitation scenario given in the deliverable D6.2. This takes into consideration the development of the first prototype and the participation of each SME in every single step of the manufacturing process, from the raw material order to the final distributed product. For each manufacturing process, the SMART-HIP component description was reminded and the main production steps were identified. When possible, an estimation was done regarding the needed time for the accomplishment of the manufacturing step and the necessary resources in term of personnel. The following component were taking into consideration (all detailed information is given in the deliverable D5.3):
Oscillator unit, developed by UMR and which will be produced by MERETE
Diagnostic device:
Extracorporeal coils, developed by UMR and which will be produced by EDE
Detection unit, developed by LABOR and which will be produced by EDE
Coils driver, developed by UMR and LABOR and which will be produced by EDE
Software and HMI, developed by LABOR and which will be maintained by EDE
Prosthesis, produced by MERETE, if applicable coated by ALHENIA
Then, the influence of the manufacturing of the SMART-HIP System on the production plant layout of each SME involved in this process is investigated. Regarding the prosthesis, MERETE and ALHENIA have already all needed equipment, production area and validated subcontractors. MERETE will also be involved in the production of the oscillator unit which will be processed through a similar production flow than any implant (assembly, packaging, sterilization). Therefore, no changes need to be done in order to implement the production of the SMART-HIP intelligent prosthesis. EDE will take over the production of the SMART-HIP Diagnostic accompanied of the software and HMI. All preliminary tests will be performed in cooperation with UMR and probably at their facilities. No any additional investment is needed to start the production of this system, as well as no investment is needed for the integration and the maintenance of the software.
The SMART-HIP system, as a medical device, needs to be certified. The following main point must be taken into account:
1.SMART-HIP intelligent prosthesis: as the SMART-HIP system is planned in a first step to be used with already existing and certified implant, no further certification is required. A certification is required only if a MERETE implant is coated by ALHENIA. Regarding the oscillator unit, UMR and LABOR took care to develop it following the standards for obtaining the CE certification. However, additional testing can be required to fulfil all specification, this point has to be investigated later after planned testing.
2.SMART-HIP Diagnostic unit and Software: EDE will try to integrate modules that are already certified as medical equipment. But some parts as the excitation coils are not certified. This will have to be done with additional testing as the Electrical Safety Testing. The software will have also to be certificated as it is related to medical devices.
Last, the costs of each component of the SMART-HIP System are estimated on the base of the manufacturing process defined. The costs include the full process from the raw material order to the final sterilization process, when the product is ready to be distributed. All detailed information is given in the deliverable D5.3.
Based on the cost analysis, a scale-up analysis is conducted, aimed at evaluating the final target price of the SMART-HIP System, both SMART-HIP Diagnostic unit with Software and SMART-HIP intelligent prosthesis.
Finally, a business plan was conducted considering that the SMART-HIP system would be integrally produced and distributed by a unique company.
WP6 – Dissemination, Exploitation and marketing
DISSEMINATION FOR THE PROJECT
The guidelines of SMART-HIP dissemination activities are based on the following principles:
1.conceive dissemination as “knowledge sharing” and bi-directional;
2.establish liaisons with industrial, research, and standardization communities;
3.transfer results to the industrial, research, and standardization communities;
4.perform close collaboration with related projects;
5.publish SMART-HIP results in relevant international scientific journals;
6.organize seminars and workshops within relevant conferences in the area, producing ad hoc brochures and posters;
7.have a web site dedicated to the project, containing both a public area, on which continuously updated information shall be published, and a restricted area for use only by the project partners.
The dissemination creates awareness and interactions between the consortium and other parties, both in industrial, academic interest groups and policy makers. The objective for the whole project was to disseminate, at the national and international level, the main results to various parties and target groups, users in particular.
The first step to spread the knowledge of the SMART-HIP project was to define and implement key activities that deliver the right messages to the right audiences using the best methods. By knowing what is useful, interesting or redundant, the project team was better able to tailor content and format to best communicate with each stakeholder group. In P1, the following analysis of the audience was conducted, confirmed for P2.
During P2, the Consortium implemented and validated the SMART-HIP dissemination strategy drafted in the interim PUDK. A consistent list of activities planned and carried out during the P2 (M10-M24) by the partners has been created and reported as a table in the Final Plan for Using and Disseminating Knowledge (D6.2).
Specific sections of the Final PUDK have been dedicated to the major dissemination activities such as:
1.The project website, running since M3 and kept updated until the end of the project;
2.The press release, brochure and poster which have been created to use as quick and visible communication tools (press release updated at M24);
3.The social networks section which shows the pages of LinkedIn, Facebook and Google+ about SMART HIP project created with the purpose of obtaining immediate publicity and spreading up the awareness on the project;
4.The link to the partners’ official pages where the project’s materials and news are reported;
5.The articles and publications created by the partners regarding the project.
In addition, relevant events, such as conferences, fairs and exhibitions and magazines/paper publication, to be possibly attended by the Consortium have been indicated with the description about title, place/link and expected date.
EXPLOITATION STRATEGY AND CONSORTIUM ROLES
Internal discussions were established during P2 regarding the possibility of each partner to take part to the exploitation of SMART HIP. This was discussed during the final meeting in Rostock on the 21-22.01.2016 and then during a Skype call on the 03.02.2016. According to what was agreed on these occasions, the SMART-HIP diagnostic device will be produced by EDE and then supply to ALHENIA and MERETE with the software.
The group of SMEs of SMART HIP will exploit the following Intellectual Property Rights gained in the project:
1.The SMART-HIP Intelligent prosthesis
2.The SMART-HIP Diagnostic system
For what concerns the SMART-HIP Intelligent Prosthesis, the system will then be commercialized directly and indirectly to the target clients by ALHENIA and MERETE, who have already reached strategic commercial agreements:
1.MERETE will manufacture and commercialize the SMART-HIP prostheses in Germany and Europe, further developing their commercial network;
2.ALHENIA with a strong international presence, will distribute the SMART-HIP prostheses throughout Europe and overseas.
3.In the medium term, MERETE and ALHENIA will also seek licensing agreements with the large manufacturers who dominate the market, to mass produce the SMART-HIP prostheses.
Regarding the SMART-HIP diagnostic device, it will be produced by EDE, and distributed by ALHENIA and MERETE in their respective markets.
The commercial strategy was discussed internally during the general meetings in P2 and the main points highlighted by the partners are reported below as a summary and confirmation of the initial strategy.
Segmentation and Targeting
Our primary market to be penetrated within 3 years after the end of the project covers clinics and private orthopaedics centres, which are more favourable to accept new technologies due to more accessible procurement processes and more open pathways to innovative products. This market is slowly gathering momentum due to increasing awareness of the need of ensuring adequate screening for prostheses failure.
The approach will be as follows:
1.We will first propose the SMART-HIP diagnostic system to the target clients for prostheses screening.
2.Following the acceptance of this methodology, ALHENIA and MERETE will propose the new SMART-HIP intelligent prosthesis, as the first prosthesis allowing implementing the new screening methodology.
After the technology has successfully penetrated the initial target market (early adopters) and reached the mature stage, we will propose it to the wider public healthcare domain and large OEM, leveraging the industrial position of ALHENIA (European and outside Europe) and MERETE (National and European) (> 3 years after project completion).
This will also open the way to further find investors to meet the extra demand, and prepare the SMEs for the full product roll out. A Porter's Five Forces analysis was conducted at the beginning of the project to determine the attractiveness of the SMART-HIP offer for these target clients and this framework emphasised that a viable business opportunity for the SMART-HIP monitoring device and prosthesis exists; 4 key success factors were identified that will help SMART-HIP compete successfully:
1.Low costs of ownership (SMART-HIP Diagnostic Device),
2.Extremely quick test execution times,
3.Extremely easy to use,
4.Drastic savings for the Healthcare systems, owing to the optimal management of the revision surgery, which prevents overuse of expensive interventions, abate the incidence of hospitalizations due to late recognition of pathological conditions.
Motivations and expectations of customers were also examined: the three most important service factors for customers in the selected clients were quality, cost, and timeliness. This supports SMART-HIP pursuing a generic niche strategy with a cost focus.
1.Positioning: SMART-HIP Unique Selling Proposition
To reach the position of leader in the intelligent prosthesis market, we need to clearly identify and propose to the target clients the SMARTHIP’s unique features, which distinguish this product from the competitors’ approach: the unique selling point of the system is to be the first affordable solution for fast and accurate detection of prosthesis, failure, and the final product will be characterized by the following features:
1.Simplicity: an easy-to-use system, characterized by the absence of time consuming, multi-step test procedures;
2.Speed: fastest testing on the market;
3.Accuracy: quick, clean and accurate;
4.Cost-effectiveness: time saving, money saving, lower cost per completed test.
SMART-HIP’s intention is to develop next-generation, integrated, point-of-care, and compact system.
2.Pricing strategy
The main pricing objectives of the SMART-HIP initiative are threefold. First, SMART-HIP must implement a pricing plan that will allow for maximization of long-term profits. This translates into a product characterized by a price premium with respect to the competitors, which justified the increase of the prosthesis production price for the introduction of the oscillator unit. Therefore, the pricing must be at a level that allows SMART-HIP to increase the SMEs market share in the target market segment, and it should be used to develop the image of the SMART-HIP brand. Then, during the full roll out (> 3 years after project completion), we will aim to reach the maximum profitability with excess capacity in order to foster the wider adoption of the system.
The product will enter on the market following different phases. First the SMART-HIP system will be introduced on the market in order to be used with MERETE prosthesis. Then MERETE could do a partnership with ALHENIA, processing the ALHENIA coating on MERETE prosthesis. Thanks to this partnership, MERETE would have the possibility to extend the SMART-HIP system on a bigger market, including ALHENIA’s contacts. As well, ALHENIA could increase his portfolio with its coating technologies. Finally, the SMART-HIP system could be sold directly to the ALHENIA customer (medical device companies and doctors), in order to be used with their own prosthesis, as the SMART-HIP system should be universal. The description of this different scenarios is given in the section 1.3 “Expected final results and their potential impact and use” of this project periodic report.
The expected impact of commercialization was summarized via a revised business plan for the 3 years following the end of SMART HIP. All details are given in the deliverable D6.2.
For what concerns the exploitable results assessed for SMART-HIP, defined during the period P1, they were finally confirmed by all partners at the end of the period P2 (M24).
IPR MANAGEMENT AND FUTURE DEVELOPMENTS
The main activity performed in P2 was the organization of internal discussions among the partners to validate or update the information drafted at M9 on the IPRs. The expected exploitable results were confirmed by all partners, and reported in the form of a table for ease of reading. We can state that, respect to what was initially proposed for the management of IPs, no relevant update was necessary. Deliverable D6.2 (Final PUDK) also shows the claims of each partner in the exploitation of these results (meaning the way that each beneficiary intends to take advantage of the results).
During the last meeting of the project, the Consortium assessed that the technology readiness level of the SMART HIP device is still not high enough to start a commercialization phase; despite the solution was validated and the proof of concept obtained as evident from the tests carried out under WP5, the SMES envisaged the necessity to proceed with additional clinical validation steps before putting the commercial plan into practice. Clinical validation will have to involve professionals and experts, foresee users’ acceptance, certification steps and in-man trials that could not be planned in the course of the project duration.
For this reason, patenting or any other protection of the generated knowledge was temporary put in stand-by in order for the Consortium to plan how to eventually prosecute with the research to address the necessary validation phases for the project. Additional funding possibilities have been explored with the Consortium, and possible calls /topics under the EU H2020 are currently under evaluation.

Potential Impact:
The final SMART-HIP system will be composed of two fundamental sub-systems:
1.The SMART-HIP Intelligent Prosthesis (EIP)
Its basic component is the Oscillator Unit, integrated in the EIP, consisting of a magnetic or magnetisable body which is fixed on a flat steel spring. The Oscillator Unit can be hosted in the implants already manufactured and certified by Merete and Alhenia, which therefore won’t need to go through an additional certification process. This oscillator enables to perform an acoustic-mechanical analysis of the variation of the resonance frequencies of total joint replacements, correlating them with the loosening status of the bone.
2.The SMART-HIP Diagnostic Device (EDD)
The proposed system for clinical use consists of the following sub-systems:
The extracorporeal coil: it is placed outside the patient’s body and used during the clinical test to excite the oscillator. The oscillator itself impinges on the joint replacement and thereby creates a sound signal (implant resonance frequencies).

The measurement device: the excitation within the implant bending modes leads to a sound emission to the surrounding bone and soft tissue, which can be detected by a vibration sensor, which is applied outside the patient. To this scope, an adequate sensor device must be developed in order to measure the produced sound signal.

The evaluation unit (software): to analyse the measured frequencies and evaluate the bony integration of the implant. The possible exploitation scenarios regarding the SMART HIP results are the following:
SMART-HIP Intelligent Prosthesis (hip stem + oscillator unit)

SMART-HIP Diagnostic device (extracorporeal coils + measurement device + HMI)

Software

From the supply chain figure reported in the DoW, it is clear that EDE will produce and supply all electronic components (oscillator unit + diagnostic unit + software) to MERETE. MERETE will produce their own certified hip stem, which can be either not coated or coated by ALHENIA. The MERETE hip stem and the oscillator unit will then be cleaned, packed and sterilized (organized by ALHENIA) and sold to the different market targets with the SMART-HIP diagnostic device and software:
MERETE can sell it everywhere in Europe (except in Spain and Portugal)

ALHENIA can sell it everywhere in the world (except in Spain, Portugal and Germany). MERETE will supply the SMART-HIP System to ALHENIA.

SFT (Other participant) is allowed to sell the SMART-HIP System in Spain and Portugal. ALHENIA will supply the SMART-HIP System to SFT.
The exploitation was discussed during the final meeting in Rostock on the 21-22.01.2016 and then during a Skype call on the 03.02.2016. According to what was discussed on these occasions, the SMART-HIP diagnostic device will be produced by EDE and then supply to ALHENIA and MERETE with the software.

Regarding SMART-HIP intelligent prosthesis (hip stem + oscillator unit), there are different possibilities:
1. The SMART-HIP diagnostic device and software are used with MERETE certified prosthesis

The MERETE prosthesis is not coated -> Scenario 1
MERETE will continue to sell its implants, accompanied from the oscillator unit, to the current byers. The SMART-HIP diagnostic device and software can be sold by MERETE, ALHENIA or SFT, according the location of the final user of the MERETE implants.
The MERETE prosthesis is coated, by ALHENIA -> Scenario 2
Involving ALHENIA in their process, MERETE can find new customer interested by special coatings and sell more implants, accompanied of the oscillator unit. The SMART-HIP diagnostic device and software can be sold by MERETE, ALHENIA or SFT, according the location of the final user of the MERETE implants.
In both case the SMART-HIP Intelligent Prosthesis (hip stem and oscillator unit, not assembled) will be cleaned and packaged by MERETE (which has already validated processes) and sterilized by a subcontractor of MERETE (BBF-Steriservice which has validated process). The SMART-HIP Intelligent Prosthesis will be mounted (hip stem and oscillator unit) during the surgery.
2.The SMART-HIP diagnostic device and software are used with certified prosthesis of other medical device companies:

MERETE customers -> Scenario 3

ALHENIA customers -> Scenario 4
In this case, MERETE and ALHENIA will continue to produce the certified implants of their customers (according the current process already performed for these customers). In addition of this implant, the oscillator unit will be provided with the implant (cleaned, packed and sterilized under the responsibility of MERETE). In both cases, the SMART-HIP diagnostic device and software can be sold by MERETE, ALHENIA or SFT, according the location of the final user of the MERETE implants.
For all scenarios:
1.EDE will provide the diagnostic unit and software as a final product to MERETE (final packaging, final labelling, and final product documentation). MERETE will provide it to ALHENIA, and ALHENIA will provide it to SFT.

2.MERETE will produce the oscillator unit and supply it as a final product to ALHENIA (final packaging, final labelling, sterilization and product documentation). ALHENIA will provide it to SFT.

List of Websites:
The website is the most versatile dissemination tool, by which the Consortium informs the broad audience about the SMART-HIP project’s purposes and objectives, its innovations and main benefits, as well as to keep them updated about the main findings achieved, performed activities and upcoming events.
The content of the web site is accessible to a general public in order to promote the exploitation of the project and research results. Each page will be subject to constant revision, and updates will be done periodically at least every 3 months. A first environment test was released on the 20th of April 2014 and the website is online since the end of M3 of the project. To access the website, please follow the link: http://www.smarthip.eu/.
In addition, the Consortium designed and created different tools for promoting the project during the dissemination activities, such as the SMART-HIP brochure and a dedicated poster. Also, a press release was implemented at the very beginning for the launch of the project and at the end of the 24 months. A screenshot of the above mentioned tools are reported in the figures below.

Final Report Summary - SMART-HIP (Sensor Based Detection of Implant Loosening in Total Hip Replacements)

Udostępnij tę stronę

Pobierz