Final Report Summary - HEAR-EU (High-resolution image-based computational inner ear modelling for surgical planning of cochlear implantation)
Executive Summary:
Just thirty-five years ago there were no effective treatments for deafness or severe hearing impairment. The advent of cochlear implants (CIs) changed that, and today implants are widely regarded as one of the great achievements of modern medicine.
However, the extent of the electrical stimulation and its effect on the brain is not well known. This, combined to the lack of pre-operative measures that predict the outcomes after implantation, results in high variability in the patient’s response after intervention. We argue that this variability could be reduced by the use of good anatomical and computational models that supply information on the part of the systems that we cannot directly observe.
The aim of the HEAR-EU project is to minimize invasiveness and insertion-induced trauma, and to improve the functional outcome of the implantation through the creation of patient-specific high-definition anatomical and functional models of the inner ear. Among the main objectives we would like to underline: a) the development of a novel high-resolution high-energy microCT device to obtain detailed images of the middle and inner ear, even in the presence of metallic implants, b) to build a model of the shape variability of the middle and inner ear from high-resolution images, also incorporating functional information, c) to build a computer-assisted patient-specific preoperative planning system d) to improve the design of cochlear implant (CI) electrode arrays and associated insertion tools using a population-based optimization framework.
Those objectives were fully reached during the course of the project, producing several important publications and exploitable results. We first concentrated on obtaining and analyzing high quality images that are used to build anatomical models (WP2). Then we provided highly detailed models of the inner ear, both on the anatomical and functional aspects. Later, we built predictive methods to estimate the precise shape of a patient’s inner ear from standard low resolution CT images (WP3). Further, in work package 4, we developed a software platform for pre- and intra-operative planning. The system allows the assessment of not only the anatomy and structures of risk, but also the possible functional outcomes of the intervention. In order to validate the final position of the implant after the surgery we developed a novel high-resolution high-energy microCT imaging system in WP5. That allows to capture accurate images of samples in the presence of metallic devices such as the cochlear implants. Finally, in WP6 we devised techniques to improve the design of current CI electrode arrays and associated insertion tools based on a population-based implant design paradigm. The statistical shape models developed in WP3 formed the basis of a computational optimization scheme that maximizes the efficacy of the electrode array on a per-patient basis.
The scientific and technological advances produced during the execution of the project have been thoroughly disseminated and their exploitation is currently ongoing by the partners of the consortium. These results will significantly improve patient’s quality of life until older age. Furthermore, they will also improve life expectations and performance for very young children allowing for even more efficient CI implantation at a very early age. Secondly, we will reduce hospitalization and surgery times required by CI surgery by making available the computerized surgical planning and diagnostic tools developed within this project. Importantly, improved functionality of CI will reduce battery consumption, again significantly lowering the running costs of this medical device. Moreover, HEAR-EU contributes to promoting the position of European industry in the EU and worldwide market of products and services associated to medical technologies.
Overall, the project has produced very important advances both theoretically and technically and the consortium is proud of what has been achieved so far.
Project Context and Objectives:
Cochlear implantation is a surgical procedure that aims to overcome hearing loss by direct electrical stimulation of the spiral ganglion cells in the inner ear. Technological progress in this area led to the development of implantable devices, which proved to be of great benefit to patients suffering from moderate to severe hearing loss.
The surgical scenario of implantation surgery is very complex. It requires high clinical expertise in order to 1) efficiently access the surgical site, the cochlea, localize nearby critical structures (e.g. facial nerve) and 2) optimize the position of the implantable device (electrode array) inside the cochlea. Furthermore, there is vast anatomical variability across patients and this makes individual optimal fitting an extremely difficult task. This also strongly influences the success of the surgery and the degree of hearing restoration. For example, within current surgical approaches it is not possible to estimate a priori the length of the cochlear duct of a patient, which varies between 25 and 35 mm. Such 30% variability in the population means that there might be a serious mismatch between the cochlear and implant frequency maps. This may lead to a confused pitch perception, reducing the benefit that could be obtained from the cochlear implant. Moreover, insertion trauma is of great concern in cochlear implantation. Hence, it is crucial that anatomical variability is considered not only during the surgical planning process, but also in the design phase of implants, in order to optimize functional outcome and reduce intracochlear damage and misplaced electrodes.
We hypothesize that a comprehensive understanding of the shape variability of the middle and inner ear among patients will be of assistance during surgical planning, and enable the design of improved hearing implants. Consequently, the aim of this project is:
1) to develop a novel high-resolution high-energy microCT device to obtain detailed images of the middle and inner ear, even in the presence of metallic implants,
2) to build a model of the shape variability of the middle and inner ear from high-resolution images, also incorporating functional information,
3) to build a computer-assisted patient-specific preoperative planning system
4) to improve the design of cochlear implant (CI) electrode arrays and associated insertion tools using a population-based optimization framework.
All objectives revolve around the criteria of minimizing invasiveness, insertion-induced trauma and enhanced functional outcome through patient-specific frequency mapping. Furthermore, due to the complexity of the surgical procedure novel technologies such as computer assisted planning could greatly benefit the training of young surgeons.
The research is based on the development of a statistical shape and functional model of the middle and inner ear, built from high-resolution cadaver microCT images. Novel high-energy microCT devices will be designed and developed. The statistical model will encapsulate the entire shape variability found in a given target population. Current techniques such as Active Shape Models (ASM) will be extended to perform automatically, on patient computer tomography (CT) and Digital Volume Tomography (DVT) scans, segmentation of the middle and inner ear, annotation of critical nearby structures of relevance for the surgical procedure, and patient-specific frequency-position mapping. This extension will be made possible by including during the modelling phase specific anatomical information (i.e. anatomical landmarks) and clinical expertise (i.e. surgical procedure), and frequency-position mapping (e.g. Greenwood functions). Additionally, the developed model will be used to optimize the design of future generations of CI electrode arrays by considering these spatio-functional and surgical requirements.
In the following we detail the objectives of the project per work package:
• The main objectives of WP1 are centered around translating the clinical requirements of the cochlear implantation procedure into detailed technical specifications for the system components to be developed. A milestone (MS1) will be reached once the clinical and technical details are clarified and the proper documentation is produced (D1.1).
• In WP2 we concentrate on obtaining and analyzing high quality images that are used to build anatomical models: acquiring the images of cadaver specimens in high resolution using micro-CT technology and designing and implementing image analysis software to identify and delineate structures of interest (segmentation) and to align pairs of images from different acquisition times or modalities for data fusion (registration). Two milestones (MS2 and MS3) will be reached once a demonstrator and report of image analysis software is available (D2.1) a parametric model of implant design is defined and documented (D6.1) and all high-resolution cadaver data are acquired, processed and documented (D2.2).
• The next package, WP3, aims at providing highly detailed models of the inner ear, both on the anatomical and on the functional aspect. Its objectives are: to develop a statistical shape model of the cochlea and surrounding structures, incorporate functional information, such as cochlear frequency maps, into the statistical model. In addition, we will build predictive methods to estimate the precise shape of a patient’s inner ear from standard low resolution CT images. To this end, a model of the modular transfer function describing the relationship between micro-CT and clinical CT will be built. Moreover, we will evaluate the predictive ability of the statistical shape model using clinical CT scans, with special emphasis on the critical structures and accuracy of the frequency map.
• Work package 4 develops a software platform for pre- and intra-operative planning, based on patient-specific models and providing decision support concerning the operability of patients, the implantation strategy and support of the pre- and intra-operative stage using appropriate visualizations. The system will allow the assessment of not only the anatomy and structures of risk, but also the possible functional outcome of the intervention. The software will be developed and designed to fit in the clinical workflow defined in WP1. The development process will be performed in close collaboration with the clinical partners at UBERN, who will participate in the validation of the system. It’s milestone, MS4, will be reached once statistical model of shape and function are created and evaluated and the demonstrator is ready (D3.1). Additionally, as the cadaver images will be scanned with the new high-energy microCT system developed by WP5, there is a dependency on D5.1.
• The development of a novel high-resolution high-energy microCT imaging system is the main objective of WP5. That would allow to capture accurate images of samples in the presence of metallic devices such as the cochlear implants. Additionally, the design and implementation of a novel software and hardware infrastructure for efficient reconstruction and pre-processing of very large image datasets will be tackled by this workpackage. Eventually, the last task is the evaluation and optimization of the high-energy microCT system to be used in the validation of the cochlear electrode optimization work of WP6. In order to complete the work and reach MS5, we will first need the complete surgical planning system constructed and tested (D4.1).
• WP6 - The main objective of this WP is to improve the design of current CI electrode arrays and associated insertion tools based on a population-based implant design paradigm. The statistical shape models developed in WP3 will form the basis of a computational optimization scheme that aims at maximizing the quality of fitting of the electrode array to the modelled population while subject to constraints derived by the surgical and functional requirements defined in WP1. The optimization framework will be generic and applicable to other types of implants and medical devices in general. This work will be considered finished and MS6 reached after the final system demonstrators is ready and evaluated (D4.2 D5.2) the implant optimization studies are finished (D6.3) and generally all milestones are achieved and the deliverables completed.
• All the management tasks are grouped in WP7, to provide the overall strategy and coordination to the project, the steering efforts of the partners for the achievement of milestones and ensuring that the work is undertaken with appropriate quality levels. Also we aim to set-up a project management structure that ensures an efficient operational management including administrative, financial and legal issues, supporting the Scientific Coordination in organizing and supervising the pilot work and in its liaison with the European Commission. Finally, we will strive here to support the appropriate communication and work dynamics to help drive the whole Consortium as a team towards successful completion.
• In WP8 we concentrate all the dissemination and exploitation efforts of the projects. We design a plan that allows for optimal communication within the project and the dissemination of information and knowledge generated by the project to relevant stakeholders. Also we disseminate the project knowledge and results to the research, clinical communities, as well as to the industry, the general public and patients. Finally, we develop an exploitation plan aiming at ensuring the use of the project results after the EC funding period.
Project Results:
By its completion at month 36, the project has successfully completed all its objectives and generated a number of important results. A summary of the main results of the different work packages is listed below:
WP1: Clinical requirements and technical system specifications
A comprehensive documentation of the most common hearing disorders, treatment approaches for cochlear implantation, and the state of the art of related surgical procedures was compiled. Clinical workflows were formalized and translated into technical specifications for the computer-assisted surgical planning system, the new high-energy microCT imaging system, and the design parameters for new cochlear implant electrode arrays.
The main results per task are:
T1.1 Clinical requirements - A literature research was performed to document the prevalence and incidence of the most common hearing disorders, treatment approaches for cochlear implantation, and the state of the art of related surgical procedures. Special attention was given to the benefits and risks of minimally invasive and robotic procedures. Clinical workflows were specified. This work was documented in deliverable D1.1
T1.2 - Technical system specifications - Technical requirements and guidelines were documented in D1.1 in relation to the design and development of the computer-assisted surgical planning system, the new high-energy microCT imaging system, and the design parameters for new cochlear implant electrode arrays. Functional system diagrams were provided for the software elements to be developed.
WP2: High resolution imaging & image processing tools
High-resolution image acquisition and processing of 44 cadaveric samples was realised, in what consists one of the largest image databases of its kind. Algorithms and software for image processing were developed and applied to these datasets. Specific methodological developments include novel image segmentation protocols and tools for CBCT and microCT data, as well as image registration methods based on surface matching with Markov Random Field regularisation, and novel algorithms for implicit registration via partial differential equations.
The main results grouped by task are:
T2.1 - High-resolution image acquisition and processing Several samples were prepared and scanned: 22 dry temporal bones - µCT (16-19 µm) 16 Thiel-fixed cochleae - both µCT (7.6 µm) and CBCT (150 µm) 6 fresh frozen cochleae - µCT (5.8-10.8 µm) 8 of the Thiel-fixed samples were rescanned because of the presence of liquid inside.
An improved µCT reconstruction algorithm was developed by SCANCO specifically to process the dataset of this study.
T2.2 – Segmentation - A literature review of the state of the art in cochlear segmentation was carried out. We devised separate strategies depending on the image modality (µCT or CBCT) and on the fixation technique (Thiel and dry). CBCT data were segmented using region growing, Otsu thresholding and level set methods Thiel-fixed data were semi-automatically segmented using a variety of tools (Seg3D, ITK-Snap, Amira). Dry data were segmented similarly to the previous case, but the basilar membrane and the lamina spiralis were not visible in those samples due to the fixation protocol and were thus not segmented.
T2.3 – Registration - We worked on 4 different types of alignment. The first one is the CBCT to CBCT registration. This is to create a statistical shape model of the overall cochlear anatomy at clinical resolution. The second method was devised to perform µCT to µCT registration and its purpose is to build the statistical shape model of the data at the highest available resolution. The third method was implicit registration via parameterization, a very promising method using partial differential equations to put in correspondence the coordinates without actually performing the registration. The last method was the creation of an atlas of dry µCT, to be used to segment new dry or clinical images.
WP3: Statistical models of inner ear shape and function
Statistical Shape Models (SSMs) have been built from both clinical CBCT scans and high-resolution microCT data. These SSMs describe in a compact manner the shape of the inner ear and the variations found in the population. A framework for patient-specific shape estimation by instantiating this model from imaging data was completed. Further, functional simulation has been incorporated into the models via Finite Element Models, to perform advanced mechanical and electrical simulations of cochlear and implant function. The software and model have been distributed for use in WP4 and WP6, from where the usability and potential are explored further.
The main results grouped by task are:
T3.1 - Statistical shape model - A pipeline to transform the segmented datasets from WP2 into a statistical shape model was created. Two statistical shape models were created, one using clinical CBCT data and the other one using microCT. In the clinical one, the point distribution model was used to create the SSM from 7 samples. The purpose of this initial model was to provide realistic deformation fields to other tasks, especially for the Finite Element Model (T3.3). The microCT-based statistical shape is built using the Statismo software and used to generate shapes corresponding to the first and second modes of variations. The first mode describes variations in the length of the cochlea, while the second mode describes variations in the width of the basal turn.
T3.2 - Patient-specific shape estimation - The shape model has been updated and finalized along with a software package for fitting the model to clinical CT images. A validation study has been performed as part of WP4. The software and model have been distributed for use in WP4 and WP6, from where the usability and potential are further explored.
T3.3 - Enrich the statistical model with functional info - This task was started ahead of planned schedule and both Alma and UPF joined this task. We created several versions of a finite element model to simulate the tonotopic properties of the cochlea. Additionally, we created an electrical model that captures the main characteristics of the stimulation of deeply implanted cochlea. The current model incorporates both mechanical and electrical simulations.
T3.4 - Functional predictive analysis - We annotated the SSM with the frequency mapping around both the spiral ganglia and the organ of Corti. We were thus able to propagate this information to both real and virtual patients. The predictions have been validated against literature.
WP4: Computer-assisted surgical planning system
A system for the CI surgical planning has been developed. The system includes the planning of patient-specific middle ear anatomy from CT/CBCT, the selection of a safe drilling trajectory to the cochlea, the definition of finer intracochlear anatomy using a statistical shape model built from microCT data and the optimization of the electrode array insertion angle for atraumatic implantation. The system is able to define safety regions to avoid critical anatomy while drilling the access to the cochlea and to use the statistical shape model of WP3 to generate customized cochlear models. Validation of all different components, as well as the integrated surgical planning system was performed.
The main results grouped by tasks are:
T4.1 - Planning system: design, development and testing - A system for CI surgical planning has been developed. The system includes the planning of patient-specific middle ear anatomy from CT/CBCT, the selection of a safe drilling trajectory to the cochlea, the definition of finer intracochlear anatomy using a statistical shape model built from microCT data, and the optimization of the electrode array insertion angle for atraumatic implantation. Furthermore, it is able to simulate the implantation of different electrode models, insertion depths, stimulation protocols, etc., in order to plan the most appropriate implantation strategy.
T4.2 - Patient-specific surgical plan - The system is able to define safety regions to avoid critical anatomy during drilling the access to the cochlea and to use the statistical shape model produced by the output of WP3 to generate customized cochlear models.
T4.3 - Evaluation of the surgical planning system - Validation was performed for all different components individually, as well as the integrated surgical planning system.
WP5: High-energy micro-CT system
The new high-energy microCT scanner prototype is up and running. The hardware development was based on a previous model (ScancoμCT-100). A new X-ray source and a new scintillator/detector system have been integrated. This new system can now be operated at a peak energy of 130 kVp, at a spatial resolution of about 6 μm and with a field of view of 13 mm. A metal artefact reduction (MAR) reconstruction algorithm has been implemented for the optimal imaging of the electrode array implanted in the cochlea. This algorithm significantly removes artefacts and improves image quality around the implant array.
A list of results grouped by tasks is below:
T5.1 - High-energy hardware design, development and testing The new high-energy microCT scanner prototype is running. The hardware development was based on a Scanco μCT 100, where a new X-ray source and a new scintillator/detector system was used. This new system can now be operated at a peak energy of 130 kVp at a spatial resolution of about 6 μm at a field of view of 13 mm.
T5.2 - Optimized software and hardware for image reconstruction and pre-processing of very large datasets - A metal artefact reduction (MAR) reconstruction algorithm has been implemented for the optimal imaging of electrode array implanted in cochlea. This algorithm significantly removes artefacts and improves image quality around the implant array.
WP6: Implant design optimization
A combination of metrics has been developed to evaluate the compliance of the array to the scala tympani size (likelihood of trauma) and the relative position between the electrodes to the stimulation targets (frequency coverage and specificity). Based on these metrics, an optimization framework was developed and applied to optimize CI implant designs. A prototype of electrode array with a new design, determined in T6.2 was manufactured. The new design was validated in a series of bench tests with plastic models of the human cochlea and cadaveric temporal bones. Furthermore, the insertion forces needed for a full insertion were also measured and compared to those for the Med-EL Flex28 electrode array. The results indicated that the new design should be as atraumatic as Flex28 itself. Subsequently, the novel high-energy microCT system, developed in WP5, was used to obtain high resolution microCT images of the implanted cadaveric cochlea with minimized metallic artefacts, to assess coupling efficiency, distance to the neural tissue, and insertion trauma. In total 4 cadaveric temporal bones were implanted. In addition to the CI electrode array development, several insertion tools were designed, constructed and tested, with the aim to assist surgeons with electrode insertion in anatomically difficult situations.
A brief summary of the results grouped by task is below:
T6.1 - Framework for model-based global optimization - The properties of the statistical shape model have been leveraged for generating a set of cochlea shapes and for automatically extract the structures of interest for frequency mapping and estimating the ideal tangential trajectory for a candidate electrode array of cochlear implant (CI).
T6.3 - Evaluation / test case - The prototype of electrode array with a new design determined in T6.2 was manufactured. The new design was validated in a series of bench tests with plastic models of the human cochlea and cadaveric temporal bones. The results indicated that the new design should be as atraumatic as Flex28 itself.
T6.4 - Framework for model-based global optimization - The properties of the statistical shape model have been leveraged for generating a set of cochlea shapes and for automatically extract the structures of interest for frequency mapping and estimating the ideal tangential trajectory for a candidate electrode array of cochlear implant (CI)
Potential Impact:
HEAR-EU has delivered a modular technology that will define and design a new generation of cochlear implants with optimal functional performance. This will significantly improve patient’s quality of life until older age, thus increasing the number of healthy life years (HEALTH work program and EIP “active and healthy aging” impacts). Furthermore, it will also improve life expectations and performance for very young children allowing for even more efficient CI implantation at a very early age. Secondly, it will reduce the costs in European health care systems (HEALTH work program impact) by a reduction of hospitalization and surgery times regarding CI surgery. This will be achieved by making available computerized surgical planning and diagnostic tools. Importantly, improved functionality of CI (reduced stimulating electrical currents associated with hearing thresholds) will reduce battery consumption, again significantly lowering the running costs of this medical device. Moreover, HEAR-EU contributes to promoting the position of European industry in the EU and worldwide market of products and services associated to medical technologies (in alignment with the European Innovation Union Initiative and in particular with the EIP “active and healthy aging”). This is a global challenge that needs to be tackled from a global European perspective to combine capacities across different countries and build synergies in view of a world-wide market.
The synergy between medical (radiology, physiology) and computer sciences is a central aspect of this proposal. It leads to a new paradigm of interdisciplinary research towards the final aim of reducing in-vivo testing by enabling in-silico simulations (in alignment with the concepts of “in-silico medicine” and the “virtual physiological human”). Furthermore, the establishment of explanatory computer models for biological and clinical processes, such as those developed in WP3, contributes to the objectives of HEALTH-2012-2 by developing tools and methods for “increasing knowledge of biological processes and mechanisms involved in normal health and in specific disease situations, to transpose this knowledge into clinical applications including disease control and treatment”.
In relation to specific impacts listed in HEALTH-2012-2.4.5-1 this project has developed: 1) highly automatized imaging tools to facilitate individual-patient oriented cochlear implant surgery and 2) innovative design of cochlear implants (topic 2.4.5-1: “refined tools, technologies and procedures aimed at helping patients with sensory impairments to improve their quality of life”). Such tools significantly enhance adequate positioning and functional fitting of cochlear implant (bionic ear) for patients with profound hearing loss. Significant improvement in cochlear implant output is expected (20% reduction in thresholds), even under more difficult hearing conditions (topic 2.4.5-1: “implantable devices and artificial organs or their parts”, and topic 2.4.5: “chronic diseases and improvement of quality of life”). Full attention is paid to “safety, bio-compatibility, interoperability and regulatory aspects” (topic 2.4.5-1) based on the expensive experience of the consortium, and in particular the industrial partners.
Further, this project also addresses the specific aspects of this call in order to encourage stronger SME efforts towards research and innovation (topic 2.4.5-1) through the involvement of two innovative SMEs, in collaboration with a large European enterprise.
The management provided the overall strategy and coordination to the project, the steering efforts of the partners for the achievement of milestones and ensuring that the work is undertaken with appropriate quality levels. Also we aim to set-up a project management structure that ensures an efficient operational management including administrative, financial and legal issues, supporting the Scientific Coordination in organising and supervising the pilot work and in its liaison with the European Commission. Finally, we will strive here to support the appropriate communication and work dynamics to help drive the whole Consortium as a team towards successful completion.
All the dissemination and exploitation efforts of the projects were concentrated in WP8. We design a plan that allows for optimal communication within the project and the dissemination of information and knowledge generated by the project to relevant stakeholders. Also we disseminate the project knowledge and results to the research, clinical communities, as well as to the industry, the general public and patients. Finally, we develop an exploitation plan aiming at ensuring the use of the project results after the EC funding period.
Dissemination activities have been carried out at all levels, from specialist scientific conferences and industrial fairs to general dissemination events and appearances in newspapers and TV. A complete exploitation plan has been submitted in D8.6.
A brief summary of the results grouped by work package is below:
T8.1 - Dissemination activities - Following the communication plan(D8.1) dissemination has followed specific strategies targeting the different audiences. Dissemination to the general public was achieved with a leaflet and appearance in several general media. Dissemination to the scientific community includes newsletters, institutional webpages, oral and poster presentations at scientific meetings and workshops. Several scientific papers have been published and also presented at international conferences. D8.3 reports the dissemination activities carried out by month 18, and D8.5 reports the dissemination activities carried out by month 36.
The strategy for communication and dissemination of the project was described in the Communication Plan (deliverable D8.1). A project website (documented in deliverable D8.2) has been created and maintained up to date. In a second iteration, the initial design of the website was modified in order to allow for more flexibility and improved aesthetic appearance. At month 18 of the project, an interim report on dissemination activities (deliverable D8.3) was submitted, giving an overview on the dissemination activities at that time. This final report is an updated version of the interim report, in which the activities of the second half of the project have been added.
Activities aiming at broad dissemination included the creation of a project leaflet, articles in newspapers and appearances in TV. Scientific articles in international conferences and journals have been produced, reporting the results of the project. Invited talks and contacts with researchers and clinical experts worldwide have been carried out.
During this project, HEAR-EU has organized three international workshops, and contributed to three more. We have actively reached out to other European research projects, and have established concrete collaborations with 5 consortia.
Companies participating in HEAR-EU have promoted awareness of the project in four industrial fairs. Policy makers and funding bodies have also been targeted, and three specific events are listed in this report, one involving the European Commission, another involving national administration in Spain, and finally an event at the United Nations headquarters in New York.
HEAR-EU has been featured twice as Success Story at the Horizon 2020 website.
T8.2 - Exploitation scenarios - The final exploitation plan (D8.6) has been submitted, indicating specific products and strategies of the partners.
Although exploitation activities are carried out and promoted by all partners of the consortium, the companies participating in the project have a leading role, for obvious reasons. In this deliverable, each of the three companies involved in the project report their intermediate exploitation plans. ALMA focuses on the development of novel surgical planning software, SCANCO reports their plans for the new high-energy micro-CT system, and MEDEL provide a business plan for the novel electrode designs that will be developed within HEAR-EU.
THE EXPLOITATION PLAN IS CONFIDENTIAL, as indicated by the classification as “Restricted” in the table of deliverables. Each company has prepared their section individually, and it has not been shared with other partners of the consortium. The project coordinator centralised the collection of the different documents and bundled them in a single document (D8.6) which is available for consultation to the EU officers.
Just thirty-five years ago there were no effective treatments for deafness or severe hearing impairment. The advent of cochlear implants (CIs) changed that, and today implants are widely regarded as one of the great achievements of modern medicine.
However, the extent of the electrical stimulation and its effect on the brain is not well known. This, combined to the lack of pre-operative measures that predict the outcomes after implantation, results in high variability in the patient’s response after intervention. We argue that this variability could be reduced by the use of good anatomical and computational models that supply information on the part of the systems that we cannot directly observe.
The aim of the HEAR-EU project is to minimize invasiveness and insertion-induced trauma, and to improve the functional outcome of the implantation through the creation of patient-specific high-definition anatomical and functional models of the inner ear. Among the main objectives we would like to underline: a) the development of a novel high-resolution high-energy microCT device to obtain detailed images of the middle and inner ear, even in the presence of metallic implants, b) to build a model of the shape variability of the middle and inner ear from high-resolution images, also incorporating functional information, c) to build a computer-assisted patient-specific preoperative planning system d) to improve the design of cochlear implant (CI) electrode arrays and associated insertion tools using a population-based optimization framework.
Those objectives were fully reached during the course of the project, producing several important publications and exploitable results. We first concentrated on obtaining and analyzing high quality images that are used to build anatomical models (WP2). Then we provided highly detailed models of the inner ear, both on the anatomical and functional aspects. Later, we built predictive methods to estimate the precise shape of a patient’s inner ear from standard low resolution CT images (WP3). Further, in work package 4, we developed a software platform for pre- and intra-operative planning. The system allows the assessment of not only the anatomy and structures of risk, but also the possible functional outcomes of the intervention. In order to validate the final position of the implant after the surgery we developed a novel high-resolution high-energy microCT imaging system in WP5. That allows to capture accurate images of samples in the presence of metallic devices such as the cochlear implants. Finally, in WP6 we devised techniques to improve the design of current CI electrode arrays and associated insertion tools based on a population-based implant design paradigm. The statistical shape models developed in WP3 formed the basis of a computational optimization scheme that maximizes the efficacy of the electrode array on a per-patient basis.
The scientific and technological advances produced during the execution of the project have been thoroughly disseminated and their exploitation is currently ongoing by the partners of the consortium. These results will significantly improve patient’s quality of life until older age. Furthermore, they will also improve life expectations and performance for very young children allowing for even more efficient CI implantation at a very early age. Secondly, we will reduce hospitalization and surgery times required by CI surgery by making available the computerized surgical planning and diagnostic tools developed within this project. Importantly, improved functionality of CI will reduce battery consumption, again significantly lowering the running costs of this medical device. Moreover, HEAR-EU contributes to promoting the position of European industry in the EU and worldwide market of products and services associated to medical technologies.
Overall, the project has produced very important advances both theoretically and technically and the consortium is proud of what has been achieved so far.
Project Context and Objectives:
Cochlear implantation is a surgical procedure that aims to overcome hearing loss by direct electrical stimulation of the spiral ganglion cells in the inner ear. Technological progress in this area led to the development of implantable devices, which proved to be of great benefit to patients suffering from moderate to severe hearing loss.
The surgical scenario of implantation surgery is very complex. It requires high clinical expertise in order to 1) efficiently access the surgical site, the cochlea, localize nearby critical structures (e.g. facial nerve) and 2) optimize the position of the implantable device (electrode array) inside the cochlea. Furthermore, there is vast anatomical variability across patients and this makes individual optimal fitting an extremely difficult task. This also strongly influences the success of the surgery and the degree of hearing restoration. For example, within current surgical approaches it is not possible to estimate a priori the length of the cochlear duct of a patient, which varies between 25 and 35 mm. Such 30% variability in the population means that there might be a serious mismatch between the cochlear and implant frequency maps. This may lead to a confused pitch perception, reducing the benefit that could be obtained from the cochlear implant. Moreover, insertion trauma is of great concern in cochlear implantation. Hence, it is crucial that anatomical variability is considered not only during the surgical planning process, but also in the design phase of implants, in order to optimize functional outcome and reduce intracochlear damage and misplaced electrodes.
We hypothesize that a comprehensive understanding of the shape variability of the middle and inner ear among patients will be of assistance during surgical planning, and enable the design of improved hearing implants. Consequently, the aim of this project is:
1) to develop a novel high-resolution high-energy microCT device to obtain detailed images of the middle and inner ear, even in the presence of metallic implants,
2) to build a model of the shape variability of the middle and inner ear from high-resolution images, also incorporating functional information,
3) to build a computer-assisted patient-specific preoperative planning system
4) to improve the design of cochlear implant (CI) electrode arrays and associated insertion tools using a population-based optimization framework.
All objectives revolve around the criteria of minimizing invasiveness, insertion-induced trauma and enhanced functional outcome through patient-specific frequency mapping. Furthermore, due to the complexity of the surgical procedure novel technologies such as computer assisted planning could greatly benefit the training of young surgeons.
The research is based on the development of a statistical shape and functional model of the middle and inner ear, built from high-resolution cadaver microCT images. Novel high-energy microCT devices will be designed and developed. The statistical model will encapsulate the entire shape variability found in a given target population. Current techniques such as Active Shape Models (ASM) will be extended to perform automatically, on patient computer tomography (CT) and Digital Volume Tomography (DVT) scans, segmentation of the middle and inner ear, annotation of critical nearby structures of relevance for the surgical procedure, and patient-specific frequency-position mapping. This extension will be made possible by including during the modelling phase specific anatomical information (i.e. anatomical landmarks) and clinical expertise (i.e. surgical procedure), and frequency-position mapping (e.g. Greenwood functions). Additionally, the developed model will be used to optimize the design of future generations of CI electrode arrays by considering these spatio-functional and surgical requirements.
In the following we detail the objectives of the project per work package:
• The main objectives of WP1 are centered around translating the clinical requirements of the cochlear implantation procedure into detailed technical specifications for the system components to be developed. A milestone (MS1) will be reached once the clinical and technical details are clarified and the proper documentation is produced (D1.1).
• In WP2 we concentrate on obtaining and analyzing high quality images that are used to build anatomical models: acquiring the images of cadaver specimens in high resolution using micro-CT technology and designing and implementing image analysis software to identify and delineate structures of interest (segmentation) and to align pairs of images from different acquisition times or modalities for data fusion (registration). Two milestones (MS2 and MS3) will be reached once a demonstrator and report of image analysis software is available (D2.1) a parametric model of implant design is defined and documented (D6.1) and all high-resolution cadaver data are acquired, processed and documented (D2.2).
• The next package, WP3, aims at providing highly detailed models of the inner ear, both on the anatomical and on the functional aspect. Its objectives are: to develop a statistical shape model of the cochlea and surrounding structures, incorporate functional information, such as cochlear frequency maps, into the statistical model. In addition, we will build predictive methods to estimate the precise shape of a patient’s inner ear from standard low resolution CT images. To this end, a model of the modular transfer function describing the relationship between micro-CT and clinical CT will be built. Moreover, we will evaluate the predictive ability of the statistical shape model using clinical CT scans, with special emphasis on the critical structures and accuracy of the frequency map.
• Work package 4 develops a software platform for pre- and intra-operative planning, based on patient-specific models and providing decision support concerning the operability of patients, the implantation strategy and support of the pre- and intra-operative stage using appropriate visualizations. The system will allow the assessment of not only the anatomy and structures of risk, but also the possible functional outcome of the intervention. The software will be developed and designed to fit in the clinical workflow defined in WP1. The development process will be performed in close collaboration with the clinical partners at UBERN, who will participate in the validation of the system. It’s milestone, MS4, will be reached once statistical model of shape and function are created and evaluated and the demonstrator is ready (D3.1). Additionally, as the cadaver images will be scanned with the new high-energy microCT system developed by WP5, there is a dependency on D5.1.
• The development of a novel high-resolution high-energy microCT imaging system is the main objective of WP5. That would allow to capture accurate images of samples in the presence of metallic devices such as the cochlear implants. Additionally, the design and implementation of a novel software and hardware infrastructure for efficient reconstruction and pre-processing of very large image datasets will be tackled by this workpackage. Eventually, the last task is the evaluation and optimization of the high-energy microCT system to be used in the validation of the cochlear electrode optimization work of WP6. In order to complete the work and reach MS5, we will first need the complete surgical planning system constructed and tested (D4.1).
• WP6 - The main objective of this WP is to improve the design of current CI electrode arrays and associated insertion tools based on a population-based implant design paradigm. The statistical shape models developed in WP3 will form the basis of a computational optimization scheme that aims at maximizing the quality of fitting of the electrode array to the modelled population while subject to constraints derived by the surgical and functional requirements defined in WP1. The optimization framework will be generic and applicable to other types of implants and medical devices in general. This work will be considered finished and MS6 reached after the final system demonstrators is ready and evaluated (D4.2 D5.2) the implant optimization studies are finished (D6.3) and generally all milestones are achieved and the deliverables completed.
• All the management tasks are grouped in WP7, to provide the overall strategy and coordination to the project, the steering efforts of the partners for the achievement of milestones and ensuring that the work is undertaken with appropriate quality levels. Also we aim to set-up a project management structure that ensures an efficient operational management including administrative, financial and legal issues, supporting the Scientific Coordination in organizing and supervising the pilot work and in its liaison with the European Commission. Finally, we will strive here to support the appropriate communication and work dynamics to help drive the whole Consortium as a team towards successful completion.
• In WP8 we concentrate all the dissemination and exploitation efforts of the projects. We design a plan that allows for optimal communication within the project and the dissemination of information and knowledge generated by the project to relevant stakeholders. Also we disseminate the project knowledge and results to the research, clinical communities, as well as to the industry, the general public and patients. Finally, we develop an exploitation plan aiming at ensuring the use of the project results after the EC funding period.
Project Results:
By its completion at month 36, the project has successfully completed all its objectives and generated a number of important results. A summary of the main results of the different work packages is listed below:
WP1: Clinical requirements and technical system specifications
A comprehensive documentation of the most common hearing disorders, treatment approaches for cochlear implantation, and the state of the art of related surgical procedures was compiled. Clinical workflows were formalized and translated into technical specifications for the computer-assisted surgical planning system, the new high-energy microCT imaging system, and the design parameters for new cochlear implant electrode arrays.
The main results per task are:
T1.1 Clinical requirements - A literature research was performed to document the prevalence and incidence of the most common hearing disorders, treatment approaches for cochlear implantation, and the state of the art of related surgical procedures. Special attention was given to the benefits and risks of minimally invasive and robotic procedures. Clinical workflows were specified. This work was documented in deliverable D1.1
T1.2 - Technical system specifications - Technical requirements and guidelines were documented in D1.1 in relation to the design and development of the computer-assisted surgical planning system, the new high-energy microCT imaging system, and the design parameters for new cochlear implant electrode arrays. Functional system diagrams were provided for the software elements to be developed.
WP2: High resolution imaging & image processing tools
High-resolution image acquisition and processing of 44 cadaveric samples was realised, in what consists one of the largest image databases of its kind. Algorithms and software for image processing were developed and applied to these datasets. Specific methodological developments include novel image segmentation protocols and tools for CBCT and microCT data, as well as image registration methods based on surface matching with Markov Random Field regularisation, and novel algorithms for implicit registration via partial differential equations.
The main results grouped by task are:
T2.1 - High-resolution image acquisition and processing Several samples were prepared and scanned: 22 dry temporal bones - µCT (16-19 µm) 16 Thiel-fixed cochleae - both µCT (7.6 µm) and CBCT (150 µm) 6 fresh frozen cochleae - µCT (5.8-10.8 µm) 8 of the Thiel-fixed samples were rescanned because of the presence of liquid inside.
An improved µCT reconstruction algorithm was developed by SCANCO specifically to process the dataset of this study.
T2.2 – Segmentation - A literature review of the state of the art in cochlear segmentation was carried out. We devised separate strategies depending on the image modality (µCT or CBCT) and on the fixation technique (Thiel and dry). CBCT data were segmented using region growing, Otsu thresholding and level set methods Thiel-fixed data were semi-automatically segmented using a variety of tools (Seg3D, ITK-Snap, Amira). Dry data were segmented similarly to the previous case, but the basilar membrane and the lamina spiralis were not visible in those samples due to the fixation protocol and were thus not segmented.
T2.3 – Registration - We worked on 4 different types of alignment. The first one is the CBCT to CBCT registration. This is to create a statistical shape model of the overall cochlear anatomy at clinical resolution. The second method was devised to perform µCT to µCT registration and its purpose is to build the statistical shape model of the data at the highest available resolution. The third method was implicit registration via parameterization, a very promising method using partial differential equations to put in correspondence the coordinates without actually performing the registration. The last method was the creation of an atlas of dry µCT, to be used to segment new dry or clinical images.
WP3: Statistical models of inner ear shape and function
Statistical Shape Models (SSMs) have been built from both clinical CBCT scans and high-resolution microCT data. These SSMs describe in a compact manner the shape of the inner ear and the variations found in the population. A framework for patient-specific shape estimation by instantiating this model from imaging data was completed. Further, functional simulation has been incorporated into the models via Finite Element Models, to perform advanced mechanical and electrical simulations of cochlear and implant function. The software and model have been distributed for use in WP4 and WP6, from where the usability and potential are explored further.
The main results grouped by task are:
T3.1 - Statistical shape model - A pipeline to transform the segmented datasets from WP2 into a statistical shape model was created. Two statistical shape models were created, one using clinical CBCT data and the other one using microCT. In the clinical one, the point distribution model was used to create the SSM from 7 samples. The purpose of this initial model was to provide realistic deformation fields to other tasks, especially for the Finite Element Model (T3.3). The microCT-based statistical shape is built using the Statismo software and used to generate shapes corresponding to the first and second modes of variations. The first mode describes variations in the length of the cochlea, while the second mode describes variations in the width of the basal turn.
T3.2 - Patient-specific shape estimation - The shape model has been updated and finalized along with a software package for fitting the model to clinical CT images. A validation study has been performed as part of WP4. The software and model have been distributed for use in WP4 and WP6, from where the usability and potential are further explored.
T3.3 - Enrich the statistical model with functional info - This task was started ahead of planned schedule and both Alma and UPF joined this task. We created several versions of a finite element model to simulate the tonotopic properties of the cochlea. Additionally, we created an electrical model that captures the main characteristics of the stimulation of deeply implanted cochlea. The current model incorporates both mechanical and electrical simulations.
T3.4 - Functional predictive analysis - We annotated the SSM with the frequency mapping around both the spiral ganglia and the organ of Corti. We were thus able to propagate this information to both real and virtual patients. The predictions have been validated against literature.
WP4: Computer-assisted surgical planning system
A system for the CI surgical planning has been developed. The system includes the planning of patient-specific middle ear anatomy from CT/CBCT, the selection of a safe drilling trajectory to the cochlea, the definition of finer intracochlear anatomy using a statistical shape model built from microCT data and the optimization of the electrode array insertion angle for atraumatic implantation. The system is able to define safety regions to avoid critical anatomy while drilling the access to the cochlea and to use the statistical shape model of WP3 to generate customized cochlear models. Validation of all different components, as well as the integrated surgical planning system was performed.
The main results grouped by tasks are:
T4.1 - Planning system: design, development and testing - A system for CI surgical planning has been developed. The system includes the planning of patient-specific middle ear anatomy from CT/CBCT, the selection of a safe drilling trajectory to the cochlea, the definition of finer intracochlear anatomy using a statistical shape model built from microCT data, and the optimization of the electrode array insertion angle for atraumatic implantation. Furthermore, it is able to simulate the implantation of different electrode models, insertion depths, stimulation protocols, etc., in order to plan the most appropriate implantation strategy.
T4.2 - Patient-specific surgical plan - The system is able to define safety regions to avoid critical anatomy during drilling the access to the cochlea and to use the statistical shape model produced by the output of WP3 to generate customized cochlear models.
T4.3 - Evaluation of the surgical planning system - Validation was performed for all different components individually, as well as the integrated surgical planning system.
WP5: High-energy micro-CT system
The new high-energy microCT scanner prototype is up and running. The hardware development was based on a previous model (ScancoμCT-100). A new X-ray source and a new scintillator/detector system have been integrated. This new system can now be operated at a peak energy of 130 kVp, at a spatial resolution of about 6 μm and with a field of view of 13 mm. A metal artefact reduction (MAR) reconstruction algorithm has been implemented for the optimal imaging of the electrode array implanted in the cochlea. This algorithm significantly removes artefacts and improves image quality around the implant array.
A list of results grouped by tasks is below:
T5.1 - High-energy hardware design, development and testing The new high-energy microCT scanner prototype is running. The hardware development was based on a Scanco μCT 100, where a new X-ray source and a new scintillator/detector system was used. This new system can now be operated at a peak energy of 130 kVp at a spatial resolution of about 6 μm at a field of view of 13 mm.
T5.2 - Optimized software and hardware for image reconstruction and pre-processing of very large datasets - A metal artefact reduction (MAR) reconstruction algorithm has been implemented for the optimal imaging of electrode array implanted in cochlea. This algorithm significantly removes artefacts and improves image quality around the implant array.
WP6: Implant design optimization
A combination of metrics has been developed to evaluate the compliance of the array to the scala tympani size (likelihood of trauma) and the relative position between the electrodes to the stimulation targets (frequency coverage and specificity). Based on these metrics, an optimization framework was developed and applied to optimize CI implant designs. A prototype of electrode array with a new design, determined in T6.2 was manufactured. The new design was validated in a series of bench tests with plastic models of the human cochlea and cadaveric temporal bones. Furthermore, the insertion forces needed for a full insertion were also measured and compared to those for the Med-EL Flex28 electrode array. The results indicated that the new design should be as atraumatic as Flex28 itself. Subsequently, the novel high-energy microCT system, developed in WP5, was used to obtain high resolution microCT images of the implanted cadaveric cochlea with minimized metallic artefacts, to assess coupling efficiency, distance to the neural tissue, and insertion trauma. In total 4 cadaveric temporal bones were implanted. In addition to the CI electrode array development, several insertion tools were designed, constructed and tested, with the aim to assist surgeons with electrode insertion in anatomically difficult situations.
A brief summary of the results grouped by task is below:
T6.1 - Framework for model-based global optimization - The properties of the statistical shape model have been leveraged for generating a set of cochlea shapes and for automatically extract the structures of interest for frequency mapping and estimating the ideal tangential trajectory for a candidate electrode array of cochlear implant (CI).
T6.3 - Evaluation / test case - The prototype of electrode array with a new design determined in T6.2 was manufactured. The new design was validated in a series of bench tests with plastic models of the human cochlea and cadaveric temporal bones. The results indicated that the new design should be as atraumatic as Flex28 itself.
T6.4 - Framework for model-based global optimization - The properties of the statistical shape model have been leveraged for generating a set of cochlea shapes and for automatically extract the structures of interest for frequency mapping and estimating the ideal tangential trajectory for a candidate electrode array of cochlear implant (CI)
Potential Impact:
HEAR-EU has delivered a modular technology that will define and design a new generation of cochlear implants with optimal functional performance. This will significantly improve patient’s quality of life until older age, thus increasing the number of healthy life years (HEALTH work program and EIP “active and healthy aging” impacts). Furthermore, it will also improve life expectations and performance for very young children allowing for even more efficient CI implantation at a very early age. Secondly, it will reduce the costs in European health care systems (HEALTH work program impact) by a reduction of hospitalization and surgery times regarding CI surgery. This will be achieved by making available computerized surgical planning and diagnostic tools. Importantly, improved functionality of CI (reduced stimulating electrical currents associated with hearing thresholds) will reduce battery consumption, again significantly lowering the running costs of this medical device. Moreover, HEAR-EU contributes to promoting the position of European industry in the EU and worldwide market of products and services associated to medical technologies (in alignment with the European Innovation Union Initiative and in particular with the EIP “active and healthy aging”). This is a global challenge that needs to be tackled from a global European perspective to combine capacities across different countries and build synergies in view of a world-wide market.
The synergy between medical (radiology, physiology) and computer sciences is a central aspect of this proposal. It leads to a new paradigm of interdisciplinary research towards the final aim of reducing in-vivo testing by enabling in-silico simulations (in alignment with the concepts of “in-silico medicine” and the “virtual physiological human”). Furthermore, the establishment of explanatory computer models for biological and clinical processes, such as those developed in WP3, contributes to the objectives of HEALTH-2012-2 by developing tools and methods for “increasing knowledge of biological processes and mechanisms involved in normal health and in specific disease situations, to transpose this knowledge into clinical applications including disease control and treatment”.
In relation to specific impacts listed in HEALTH-2012-2.4.5-1 this project has developed: 1) highly automatized imaging tools to facilitate individual-patient oriented cochlear implant surgery and 2) innovative design of cochlear implants (topic 2.4.5-1: “refined tools, technologies and procedures aimed at helping patients with sensory impairments to improve their quality of life”). Such tools significantly enhance adequate positioning and functional fitting of cochlear implant (bionic ear) for patients with profound hearing loss. Significant improvement in cochlear implant output is expected (20% reduction in thresholds), even under more difficult hearing conditions (topic 2.4.5-1: “implantable devices and artificial organs or their parts”, and topic 2.4.5: “chronic diseases and improvement of quality of life”). Full attention is paid to “safety, bio-compatibility, interoperability and regulatory aspects” (topic 2.4.5-1) based on the expensive experience of the consortium, and in particular the industrial partners.
Further, this project also addresses the specific aspects of this call in order to encourage stronger SME efforts towards research and innovation (topic 2.4.5-1) through the involvement of two innovative SMEs, in collaboration with a large European enterprise.
The management provided the overall strategy and coordination to the project, the steering efforts of the partners for the achievement of milestones and ensuring that the work is undertaken with appropriate quality levels. Also we aim to set-up a project management structure that ensures an efficient operational management including administrative, financial and legal issues, supporting the Scientific Coordination in organising and supervising the pilot work and in its liaison with the European Commission. Finally, we will strive here to support the appropriate communication and work dynamics to help drive the whole Consortium as a team towards successful completion.
All the dissemination and exploitation efforts of the projects were concentrated in WP8. We design a plan that allows for optimal communication within the project and the dissemination of information and knowledge generated by the project to relevant stakeholders. Also we disseminate the project knowledge and results to the research, clinical communities, as well as to the industry, the general public and patients. Finally, we develop an exploitation plan aiming at ensuring the use of the project results after the EC funding period.
Dissemination activities have been carried out at all levels, from specialist scientific conferences and industrial fairs to general dissemination events and appearances in newspapers and TV. A complete exploitation plan has been submitted in D8.6.
A brief summary of the results grouped by work package is below:
T8.1 - Dissemination activities - Following the communication plan(D8.1) dissemination has followed specific strategies targeting the different audiences. Dissemination to the general public was achieved with a leaflet and appearance in several general media. Dissemination to the scientific community includes newsletters, institutional webpages, oral and poster presentations at scientific meetings and workshops. Several scientific papers have been published and also presented at international conferences. D8.3 reports the dissemination activities carried out by month 18, and D8.5 reports the dissemination activities carried out by month 36.
The strategy for communication and dissemination of the project was described in the Communication Plan (deliverable D8.1). A project website (documented in deliverable D8.2) has been created and maintained up to date. In a second iteration, the initial design of the website was modified in order to allow for more flexibility and improved aesthetic appearance. At month 18 of the project, an interim report on dissemination activities (deliverable D8.3) was submitted, giving an overview on the dissemination activities at that time. This final report is an updated version of the interim report, in which the activities of the second half of the project have been added.
Activities aiming at broad dissemination included the creation of a project leaflet, articles in newspapers and appearances in TV. Scientific articles in international conferences and journals have been produced, reporting the results of the project. Invited talks and contacts with researchers and clinical experts worldwide have been carried out.
During this project, HEAR-EU has organized three international workshops, and contributed to three more. We have actively reached out to other European research projects, and have established concrete collaborations with 5 consortia.
Companies participating in HEAR-EU have promoted awareness of the project in four industrial fairs. Policy makers and funding bodies have also been targeted, and three specific events are listed in this report, one involving the European Commission, another involving national administration in Spain, and finally an event at the United Nations headquarters in New York.
HEAR-EU has been featured twice as Success Story at the Horizon 2020 website.
T8.2 - Exploitation scenarios - The final exploitation plan (D8.6) has been submitted, indicating specific products and strategies of the partners.
Although exploitation activities are carried out and promoted by all partners of the consortium, the companies participating in the project have a leading role, for obvious reasons. In this deliverable, each of the three companies involved in the project report their intermediate exploitation plans. ALMA focuses on the development of novel surgical planning software, SCANCO reports their plans for the new high-energy micro-CT system, and MEDEL provide a business plan for the novel electrode designs that will be developed within HEAR-EU.
THE EXPLOITATION PLAN IS CONFIDENTIAL, as indicated by the classification as “Restricted” in the table of deliverables. Each company has prepared their section individually, and it has not been shared with other partners of the consortium. The project coordinator centralised the collection of the different documents and bundled them in a single document (D8.6) which is available for consultation to the EU officers.
