Breast biopsy system guided by Positron Emission Mammography allowing real-time 3D visualization of tumour lesion and needle insertion guidance for higher sampling accuracy and efficiency

Final Report Summary - MAMMOCARE (Breast biopsy system guided by Positron Emission Mammography allowing real-time 3D visualization of tumour lesion and needle insertion guidance for higher sampling accuracy and efficiency)

Executive Summary:
Breast cancer is the most frequent cancer among women and one of the leading causes of cancer-related mortality and early diagnosis is essential to reduce the risk of mortality. First diagnostic tool is clinical imaging (mainly Mammography and Ultrasound) but suspicious findings usually require a biopsy to confirm the diagnosis; 2% of women who undergo a screening mammogram will require some type of breast biopsy (1,5 million women each year in Europe).

Breast biopsy is an image-guided procedure that can use different imaging technologies. Apart from US-guided biopsy, which is a manual procedure with limited indications, all current solutions are based on discrete scans of static images taken at different steps during the procedure. This guidance method makes current biopsy techniques to be long procedures with limited accuracy.

MAMMOCARE project has developed a PEM (Positron Emission Mammography)-guided breast biopsy system allowing real-time 3D visualization of the lesion and real-time guidance with continuous monitoring of both, lesion position and needle motion. The system allows for preoperative planning of the optimal needle path. Exclusive PEM technology used, having the highest sensitivity and spatial resolution in the market, allows the detection of small lesions difficult to find with conventional morphological imaging technologies, contributing to earlier diagnosis.

The consortium integrates 3 SMEs with experience in PEM imaging technology, Mechatronics applied to medical devices and Software Development: ONCOVISION (Coordinator) (Spain), Statice (France) and Virtual Angle (Netherlands). Oncovision is a Spanish technologic company, specialized in Molecular Vision applied to Health Sciences focusing in: cancer diagnosis and treatment, and advanced research in neurology, oncology and cardiology. Statice is an innovative SME, which main activity is providing R&D services in intelligent Microsystems in Mechatronics and in Biomaterials. Its main markets are medical, in vitro diagnostics, non-destructive tests. Virtual Angle provides solutions, services, and technologies for mission and business critical information systems. VA supports customers across several markets including telecom, medical, clinical, public sector, industry, aerospace, and defense.

Research & Development activities have been mainly carried out by the Institute of Instrumentation in Molecular Imaging (I3M) of the Spanish National Research Council (CSIC) (Spain), the Instituto de Biomecánica de Valencia (IBV – Biomechanics Institute of Valencia) (Spain) and the UK Health & Environment Research Institute (UK).

Clinical feasibility of the system has been assessed through a clinical validation study at the facilities of Stichting Het Nederlands Kanker Instituut – Antoni Van Leeuwenhoek Ziekenhuis (Netherlands), collaborating end-user, therefore demonstrating the MAMMOCARE contribution to defeat women breast cancer.

Project Context and Objectives:
Breast cancer is by far the most frequent cancer among women, with an estimated 1,5 million new cancer cases diagnosed in 2012 and one of the leading causes of cancer-related mortality. Just within the European Union, every 2.5 minutes a woman is diagnosed with breast cancer, and every 7.5 minutes a woman dies from the disease (3% of cancer deaths in European women is produced by breast cancer). Besides, it is estimated that about one in 12 women will develop the disease before the age of 75 years.

As with most other types of cancer, breast cancer is more easily treatable when it is identified early, before it has had the chance to spread to other tissues. It is demonstrated that early diagnosis is essential to reduce the risk of mortality, since 5-year survival rate is approximately 99% for patients diagnosed with local breast cancer, 84% for those with regional disease, but only 23% for patients with distant metastases.

All women with breast abnormalities are assessed using the triple diagnostic method, and the diagnosis accuracy of breast lesions depends on the correlation of clinical findings, imaging features of a breast lesion and the results of the biopsy.

Imaging is not enough to assess whether a suspicious finding represents breast cancer, another type of pathology, or normal tissue. Although mammography (which is the gold standard) is important for the early recognition of disease, less than 1% of patients with a suspicious mammogram actually have cancer. Therefore, most abnormal mammograms are false-positive findings that require additional evaluation. Roughly 10% of women who undergo a screening mammogram will require some type of follow-up evaluation, with approximately 2% requiring some type of breast biopsy, being around 1,3 million breast biopsies performed each year in Europe, which are the potential beneficiaries from MAMMOCARE project.

The objective of a breast biopsy is to obtain cells for microscopic evaluation from a suspicious breast lesion. Examination of biopsy samples by a pathologist is essential in diagnosing suspicious breast masses, determining how far the patient's cancer has advanced, and deciding on a course of treatment.

Biopsy of breast lesions is increasingly being performed with a variety of image-guided techniques, which has led to a steady decline in the number of diagnostic open surgical biopsies performed over the years. No biopsy method is 100% accurate. The most important measure of biopsy accuracy is the false-negative rate. In a study population, the false-negative rate is the percentage of malignant tumours diagnosed as completely benign (exclusive of high-risk lesions) at initial biopsy and later proved malignant.

Image-guided biopsy can be done under guidance of various conventional imaging technologies including X-ray stereotactic, Ultrasound and MRI, having each technique its inherent accuracy and specificity. Last trends in research for breast cancer diagnosis have demonstrated the potential of Positron Emission Mammography (PEM) functional imaging, as a complementary diagnosis tool and especially indicated for women in whom evaluation and diagnosis with standard methods like X-ray mammography is more difficult (ie. cystic or radiographically dense breasts). There is only one PEM-guided biopsy system on the market. It is the PEM-FLEX system developed by the American company NAVISCAN. The system’s design is based on a standard mammogram design, but using PET imaging instead of X-ray imaging. Despite being able to detect smaller lesions, clinical studies demonstrate 8% of the biopsies require an additional biopsy because of errors mainly related with bad needle positioning and movement of the lesion, highlighting lack of accuracy for targeting the lesion. Furthermore, the biopsy process is considerably long (up to 120 minutes), which is being a real commercial disadvantage.

Apart from ultrasound guidance, a manual procedure, which has its inherent imaging limitations and the biopsy accuracy relies on the practitioner’s experience and is influenced by patient movements, all current image-guided biopsy solutions are based on discrete scans of static images taken at different steps during the procedure. Furthermore, after sample extraction, if a biopsy error is detected in the post-biopsy scan, the whole biopsy procedure should be repeated.

To sum up, current biopsy guiding technologies are long procedures with limited accuracy.

Guidance based on static preoperative image is inaccurate, as deformation and movement of soft tissue during biopsy takes place, which may lead to needle positioning errors and inadequate sampling because of movements of the target lesion during needle penetration. Real-time monitoring of both, the lesion position and needle motion during the whole biopsy procedure would assure adequate sampling of the lesion. Real-time monitoring will also allow for continuous checking of the needle insertion process and the possibility to correct it in case of detected deviation.

It is possible to miss a lesion all together or to get a sample out of tissue target region. Apart from the serious consequences of an inadequate tissue sampling, which may lead to false negatives or wrong diagnoses, positioning errors may lead to more than one perforation of the breast, which implies more discomfort for the patient and a higher risk of pain and infection after the procedure. Hence, the higher the accuracy of the biopsy positioning system, the lower the extraction errors and the higher the ability to locate and extract small lesions like the ones detected by PEM imaging technology (1,5-2 mm).

These findings, together with the growing availability of PEM technique in specialized cancer centres, as well as the growing technological capabilities in processing, storing and treating clinical images, specially PEM, provides a unique opportunity to address a need in a market that, currently, has not been met: early diagnosis of small breast cancers, with high accessibility, and be able to carry out the biopsy procedure through real time visualization, continuously monitoring the needle trajectory with respect to the lesion; and achieving the necessary accuracy thanks to high precision mechanisms.

MAMMOCARE project has developed a breast biopsy system guided by 3D real-time PEM (Positron Emission Mammography) imaging and offering real-time guidance and correction of the needle trajectory for higher accuracy and efficiency in tumour sampling. It is be the first breast biopsy system in the market offering real-time monitoring and guidance. The system automatically calculates the best needle position to perform the biopsy and it is placed and inserted accordingly for the physician to extract the selected sample.

Exclusive PEM image technology used in the system, having the highest spatial resolution and sensitivity in the market, allows for detection of smaller lesions (1,5-2 mm) that could not be found out with other more conventional image technologies like RX, US or MRI, or even other PEM scans in the market, and is also able to display heterogeneous uptake in malignant breast tumours, contributing to earlier and more precise diagnoses.

Biopsy module allows for real-time guidance and positioning of the needle, with higher accuracy than current image-guided biopsy systems in order to enable targeted sampling of small tumours, thus making feasible to diagnose breast cancer at very early stages, improving patient prognosis and survival chances. In addition, the biopsy module is more accessible in order to reach more difficult areas not possible with other techniques, while reducing risk. In addition, procedure duration is shorter than the current PEM-guided biopsy, offering overall a cost-efficient alternative for breast cancer diagnosis.

The MAMMOCARE system presents the following features:
• Real-time visualization of 3D PEM images: real-time visualization of the lesion’s spatial position before and during the biopsy process.
• The system is based on proprietary PEM technology with the highest sensitivity and specificity in the market, able to detect small breast malignancies (of the order of 1.5 mm) and to differentiate between different zones of the tumour according to their cellular/chemical activity (this allowing for an optimal selection of the area of interest for sample extraction). PEM is the only imaging technology allowing direct detection of breast cancer cells.
• Biopsy System with precision mechanisms allowing a high spatial accuracy in the needle positioning and accurate biopsy of the small tumour structures detected with PEM technology.
• Smart biopsy control system: automatic calculation of the 3D position of the target lesion selected by the physician; planning of the optimum and least invasive needle path to perform the biopsy; and continuous monitoring of the biopsy procedure through a navigation software able to detect any misalignment of the needle trajectory.
• Rotary compression plates with pressure measuring sensors. This is a unique feature that will allow immobilizing the breast according to previous calculations of the optimum approach, and subsequent needle positioning. The plates are sensorized so that the breast is compressed to an optimum pressure degree: Mild compression of the breast during the whole biopsy procedure (less pain and more comfort for the patient), minimizing tissue deformation during the needle insertion.
• Real-time correction of the needle trajectory is suggested in the navigation software if lesion moves from the planned trajectory during the procedure over a confidence threshold. Lesion, needle and planned trajectory will be seen real-time on the screen. Suggested correction will be displayed in order to ensure adequate positioning with regard to the lesion, by physician’s re-selection of the moved target lesion, thus avoiding the need for more than one breast perforation.
• High accessibility thanks to the prone position configuration (more comfortable for the patient), together with the rotary compression plates, which allows the biopsy of those regions with more difficult access, like those near thoracic wall or in the superior external breast quadrant. This feature increases the number of lesions detected with respect to the competence.
• Compatible with different commercial vacuum-assisted biopsy instrumentation.

MAMMOCARE produces the following main benefits:

• Earlier cancer diagnosis with the proprietary PEM technology, since it distinguishes between an active and a non-active cancer tissue that may occur before there is any change in gross anatomy (which could then be seen on conventional morphological imaging techniques).
• The ability to differentiate vital parts of the tumour using the dedicated PET imaging technology allows obtaining biopsies from the most active and metabolic areas of the tumours, which may lead to a more accurate tumour sampling and improved tumour tissue characterization (improved genomic profiling).
• 3D real-time PEM-guided biopsy leads to a shorter and innovative procedure (30-40 minutes) based on continuous checking of the lesion position and needle insertion process.
• Comfort and low risk, thanks to the smart biopsy control system together with the prone position and the rotary compression plates, which allows the optimal approach for the biopsy procedure.
• Better accuracy of the biopsy procedure (real-time guiding accuracy of ± 0.5 mm), according to PEM imaging performance, minimizing errors in needle positioning and sample extraction and avoiding the need for a second biopsy procedure or more than one perforation of the breast.
• The higher reliability in tumour sampling allows for better treatment definition and planning, minimizing the side effects for the patient (reducing the associated costs due to radiotherapy or chemotherapy sessions).
• PEM technology, apart from diagnosis, allows for the later therapy treatment follow-up.
• Therefore, earlier diagnosis provided with this technique will allow savings for national Public Health Systems (savings around 7,6 M€ per year just because of the reduction of procedure duration).
• Improvement of the quality of life of breast cancer patients and higher patient survival expectancies

Project Results:
The main result of the project is the development of a Mammocare prototype, a breast biopsy system guided by 3D real-time PEM (Positron Emission Mammography) imaging and offering real-time guidance and correction of the needle trajectory for higher accuracy and efficiency in tumour sampling.

During the whole duration of the project, all the subsystems have been completely developed:

1. PEM imaging module
2. Biopsy module
3. Biopsy monitoring and control software, including the Graphical User Interface

All these subsystems were integrated into a final prototype. The main structure that was designed to give support to all subsystems was based on a rotation plate. This Rotation plate was designed for support the PET detector, the PET open and close mechanisms, the compression paddles and the biopsy subsystem. The Mammocare system was calibrated and in-vitro validation tests were conducted before sending it to NKI-AVL’s facilities.

The prototype developed has reached the following highlights:

1. Patient and User ergonomy:
The final prototype configuration is a compact device with similar configuration to standard biopsy stereotaxia systems on the market.
From the point of view of the patient, MAMMOCARE offers a confortable system, adjustable bed and curved compression paddles.
From the point of view of the clinician, MAMMOCARE offers a free open system for the user, were the bed height can be adjusted. The visualizing system is close to the patient, were the status of the biopsy procedure can be followed.
Figure 1. MAMMOCARE installed on NKI

2. Biopsy compatible PET system:
The new PET detector system has been designed with an operational open ring configuration during biopsy procedure. This open/close configurable system provides the best specifications (sensitivity and spatial resolution) during standard scan (closed ring) and active PET detector with a 60mm open window.
Figure 2. Double ring detector with detailed mechanics for open and closed procedure

3. Mechatronic system compatible with biopsy needle:
The final mechatronics is compatible with eVIVA needle and preserves the sterility of the setup with a laser calibration of the setup.
The system moves automatically into the shortest way to the lesion. The technician commands the final puncture manually.
Figure 3. Needle mechatronics system
A support for the hydromarker insertion tool has been added, giving the capability to the system to insert a hydromarker in the biopsied breast zone.

4. User software:
The software provides the user with the flexibility to proceed to biopsy into different protocols, keeping the control about positon of tissue, position of needle and breast fixing preasure.

The system has been installed at NKI-AVL’s facilities for its calibration and validation. A clinical study was conducted with several patients showing positive results.

The following pages describe the work performed at WorkPackage level.

Work Package 1. PEM Imaging Module Development.

In this work package, the consortium has designed and developed a PEM imaging module. The new detector designed has better specifications than originally described on the DoW. The final detector version is a double ring detector with a motorized system for open a close the ring offering a active mode with a 60mm gap and a maintenance/patient operation mode with 100mm gap.

Each detector ring is configured by 12 single detector modules. Each detector module has a simple configuration based on:

- Monolithic LYSO crystals.
- Hamamatsu Position Sensitive Photomultiplier Tube (PSPMTs)
- Power and amplification electronics

Figure 4. Module assembling distribution

The detector developed is a double ring PET, which can be separated in two symmetrical semi-rings and being an operative detector on this configuration

Figure 5 CAD detector design. Figure 6 final PET detector configuration

a) Semi-ring open/close system
The biopsy operation can start only when the PET ring is open:

Figure 7. Open ring position

This configuration of the PET system is new on the market, and defines a different detector configuration that allows a 60mm window (on the more limitant zone) for needle corrections. It can be opened up to a window of 100mm in order to enhance clinician operation during first incision and paddle removal.
Figure 8. Open ring detail

The semi-ring open/close system is currently finished and its prototype is being manufactured as the breast immobilisation system.
The semi-rings will be mounted directly in a spindle unit. We use an intermediate piece to align the top of the rings with the top of the paddle (Figure 9).

Figure 9. Opened and closed semi-rings

The spindle unit has two carriages and the thread of the spindle is both right and left-hand. The rings can be opened or closed symmetrically (Figure 10).

Figure 10: Spindle unit

The system is operated with a stepping motor connected to the spindle with a belt (Figure 11).

Figure 11: System detail

Electronic design.
The electronic design was based on known MAMMI technology including some modifications to make it compatible with new requirements of MAMMOCARE. The acquisition system is based on a modular architecture with expandable clock, centralized in one control card shot. This card is responsible for generating the timing signals of all the elements and controlling the start of conversion of each of the A/D cards according of a number of parameters that have been previously scheduled.

Figure 12. Ring electronics configuration

When a gamma ray hits two opposite detectors, inside a time window, according to a predefined combination of detectors, the trigger card detects a valid event.
The Analog to Digital Converter (ADC) cards send the data through 100 Mbps Ethernet ports. These ports are connected to a 24-input/1-output switch for data transmission, computer storage and processing.
All control and digital processing technologies are based on CPLD logic and programmable FPGA.
These devices implement parallelizable algorithms that work in real time and are designed to significantly increase performance and speed over microprocessor-based systems.
It consists on a card shot (trigger) for up to 24 entries and a maximum of 6 ADC cards interconnected through a rear panel (backplane). Each card has 24 acquisition analog inputs from 4 detector modules and 2 digital outputs. This card can’t operate autonomously. It has to be connected to the backplane trigger card, as this is responsible for generating the clock signals and necessary control for operation.
The system must be mounted in a cPCI format rack. The power supply must be independent to avoid noise from other elements.

The following figure shows final electronics configuration on final setup versus the initial design.

Figure 13 Schematic detector cabling. Figure 14 final detector electronics distribution

PEM imaging reconstruction algorithms

The algorithms commonly used for reconstruction is the MLEM. This iterative algorithm works as follows. Starting with a uniform image (all voxels are set to one), it projects the image into the LORs space, to get the set of LORs that should generate that image (this step is called forward projection). Then it compares that set of LOR with the actual LORs detected (histogram or sinogram), applying the error factor for each LOR. The last step is to update the image with that new information (this step is called backprojection), and then it repeats the forward and back projection until the condition of termination has been reached.
All the calculation performed in the forward and back projections need a probability matrix, which models the system. These probabilities are the key of the algorithm and they need to be calculated as accurate as possible modeling the current system.
In this task, the research group has developed a method of calculating these probabilities for the MAMMOCARE project in all the geometric configurations needed.
1. Probability Calculation Method. The selected method has been based on the Solid Angle (An object's solid angle is equal to the area of the segment of a unit sphere, centered at the angle's vertex, that the object covers).
The calculation has been performed projecting each virtual pixel in the detector through all the possible voxels and calculating the intersection area with all the pixels at the opposite detector. The solid angle of that area (always inside a pixel) has been used as the probability estimator.

Figure 15. Pixel projection.

2. Closed Ring Configuration
A complete software for the calculation of probabilities has been generated, allowing the setup of parameters as virtual pixels in the detectors, geometric distances, module number, etc...

Figure 16. Calculation software. Closed ring.
The closed ring geometry probabilities have been calculated using the solid angle estimator, generating all the possible LOR and Voxel combinations for a virtual pixelation of 20x20 pixels per detector.

3. Open Ring Configuration
The previous software has been modified to allow the calculations for the open ring configuration geometry.

Figure 17. Calculation software. Open ring.
The open ring geometry probabilities have been calculated using the solid angle estimator, generating all the possible LOR and Voxel combinations for a virtual pixelation of 20x20 pixels per detector.

Software for real-time reconstruction and 3D imaging of the breast
The main part of this work is based on the fast reconstruction algorithms developed on T1.3 which use List Mode (LM) algorithms. As their name suggests, they use the list mode format for reconstruction, implying that the data does not need to be histogrammed.

Currently, the LM-OS algorithm implemented in Mammocare was developed based on an overhead Siddon method [6], well known as a computationally efficient mechanism. Siddon considers the probability of a voxel emission in a defined LOR is directly related to the LOR segment traversing the same voxel. This approach is extremely fast, but the image quality obtained is not comparable to standard algorithms since they model the probabilities with more approximate methods.

The final reconstruction time obtained using this method was around 12sec, fast enough for real time imaging.

These reconstruction algorithms were adapted to be compatible with the GUI software.

Characterization of the PEM imaging module
For a complete proof of concept validation of all the detector subsystem the development team needed to perform a characterization test of the detector on the 2 active configurations: open and closed ring. This first step provided relevant information about what will be expected to have with open ring configuration.

For closed ring configuration, the team followed MAMMI acceptance criteria, as the detector has the same core design and must have the same performance.

Final specifications for the double function ring detector.

Spatial resolution:

- Typical: 1.8 mm
- Acceptance Criteria: <2.1 mm
- Results Close Ring:
o X Axis: 1.52 mm
o Y Axis: 1.55 mm
- Results Open Ring:
o X Axis: 1.71 mm
o Y Axis: 2.44 mm

Sensitivity

- Typical: Center ring
- Acceptance Criteria : 1.3%
- Results Close Ring: >1.2%
- Results Open Ring: >1.2%

Uniformity

- Typical: 10%
- Acceptance Criteria : <20%
- Results Close Ring:
o Measured Integral Total: 12.60 %
o Measured Integral Central: 11.75 %
o Measured Differential Total: 8.13 %
o Measured Differential Central: 8.13 %
- Results Open Ring:
o Measured Integral Total: 30.83 %
o Measured Integral Central: 27.17 %
o Measured Differential Total: 16.79 %
o Measured Differential Central: 12.09 %

Energy resolution

- Typical: 20%
- Acceptance Criteria : <30%
- Results Close Ring: 25.07 %
- Results Open Ring: 25.07 %

We found a reduction of spatial resolution and uniformity of the system on open mode as expected, being operative for biopsy functions.

On a second step of this proof of concept, the team matched some object acquisition with open ring and closed one in order make a final cross check between both detection systems.

Figure 18 Closed ring cross correlation. Figure 19 Open ring close correlation

In conclusion, the new detector developed for the MAMMOCARE prototype preserves performance of MAMMI design and provides a open design to perform biopsies.

The final performance obtained with open ring detector is similar to the closed ring one, with an impact basically on x-axis with bigger spatial resolution, remaining unaffected the capability to see a lesion.

Work Package 2. Biopsy Module Development

The consortium has developed a computer-guided biopsy positioning system, with the following specifications:

• Fully compatible with the PEM imaging system developed in WP1.
• With an accurate control of biopsy needle positioning.
• Able to calculate and correct the optimal path for the needle based on real-time calculation of 3D position of the
lesion and the needle itself.
• Compatible with current vacuum-assisted breast biopsy systems in the market

Biopsy system’s requirements
In this first task, the biopsy system requirements were defined. The mechanical parameters were specified, this included the number of axes and the range of movement that the biopsy positioning system is required to move in order the position the biopsy needle correctly in relation to the region of interest within the breast tissue. A high level flow chart showing the functionality of the biopsy system was developed and the following points were considered:
• Operating range
• Accuracy/resolution
• Referencing/Datum point
• Operational Speeds/Time
• Forces
• Emergency override
• Biopsy Device Attachment
• Regulatory requirements
• Materials
• Cleaning/Sterilising

The electrical and control modules that are required for operating the biopsy positioning system were also specified. The specification is based on the equipment being compatible with the other elements of the overall Mammocare system. This is important for integration.
The outcome of the work in task 2.1 was the biopsy system requirements specification, which formed the D2.1 deliverable report.

Mechanical design of the biopsy system
The biopsy positioning system was designed in conjunction with the designs for the breast immobilisation system and the PEM positioning system to ensure that all elements operate within the constraints imposed by each system. The design of biopsy positioning system has been based on an Altazimuth mount system using two rotational axes mounted perpendicular to each other i.e. as typically used for mounting cameras, telescopes, radio antennas, gun mounts etc. Rotation about the vertical axis moves the azimuth (compass bearing) of the biopsy device while rotation around the horizontal axis alters the angle of elevation (altitude) of the biopsy device enabling the biopsy needle tip to be targeted at any point within the PEM scan volume.

To achieve accurate movement of the azimuth and altitude axes precision worm gear driven rotary axes that have been used are capable of achieving micron levels of accuracy. Movement in the insertion direction of the needle is via a lead screw driven linear actuator, initial movement up to edge of the PEM scan area is powered using a stepper motor. As the insertion of the needle is patient critical this part of the movement is manually controlled by the clinician via a hand wheel. To achieve the required positional control of each axis encoders have been incorporated into each drive which communicate with the biopsy system control functions. Additionally absolute encoders have been incorporated into each axis to provide real time position data on the actual position of the axis for use in position confirmation and to feed back to the operator interface to provide a visual indication of the control screen.

Figure 20. Biopsy needle positioning system

Design of the breast immobilization system
The breast immobilisation system was designed, and its prototype manufactured. This prototype was assembled to the other sub-systems: biopsy positioning system, detection system …

The breast immobilisation system consists of two paddles, which compress and fasten the breast to allow a proper biopsy (Figure 21).

The paddles were mounted into a spindle unit to open and close the paddles. The spindle unit is operated by a pulley from the bottom of the rotator plate

Figure 21. Breast immobilization system

The paddle is made of polycarbonate with rounded edges and with a window to access the breast and to perform the biopsy. In this part, there is a reservoir to collect blood and a protrusion to allow an easy removing of this part from the system in order to clean it.
The clamping force could arrive to 20 kg but physicians usually compress the breast less than 11 kg. It produces a usual bending moment lower than 20 N.m and a maximum bending moment around 30 N.m (Figure 22).
In order to measure the clamping force of the paddle, we use a load cell; and we place the system into a ball bearing guide to avoid the bending moment in the load cell (Figure 23).

Figure 22. Paddle

Figure 23. Load cell mounting

The paddle system is mounted in a spindle linear unit. Due to the high bending moment of the paddle, we need a robust linear unit. The spindle unit has two carriages, the thread of the spindle is both right- and left-hand; so you can open or close the paddles symmetrically (Figure 24).

Figure 24: Spindle unit

The system can be operated by a stepping motor or manually with a wheel if the clinicians want to adjust manually the compression force or to open the paddles manually in case of system failure. The torque is transmitted by a belt from the angular drive to the spindle unit. We use an angular drive to connect both motor and manual system (Figure 25).

To connect the motor to the angular drive, we use an elastic coupling that allows axial, lateral and angular misalignments.

We have placed the motor, the wheels and the angular drive below the rotator plate in order to have space enough for the biopsy mechatronic guidance system. (The distance between the front of the rings and the pulley is around 13 centimetres).

Figure 25: Paddle drive system

Figure 26: Wheel Figure 27: Angular drive

To prevent a high compression load in the breast we use a torque limiter between the pulley and the spindle unit (Figure 31). We can adjust the maximum torque and then, the maximum compression force of the paddles. We have chosen a torque range interval from 1 Nm to 3 Nm.

Figure 28: Torque limiter

Figure 29. Prototype assembling (I)

Figure 30. Prototype assembling (II)

Electronic control unit for needle motion
Based on the design of the PEM ring and the space available for the movement of the biopsy needle, the detailed design of the mechanics was undertaken. The compatibility with the PEM was an important factor that was considered.
The selection of precisions components was investigated and the components that were finally selected were designed into the prototype.

For this task, HERI has provided a parts list, information and block diagrams.
A schematic diagram with the electronic modules was produced and the following components were selected for the MAMMOCARE application:
SSI Board with BISS, processor Board, RS485 4 port Cpci, PCI subrack with PSU, Laser displacement sensor, Stepper motor with integrated controller, encoders.

Software for control of the needle motion
As a first step, the software block diagram and flow chart were developed and the method to be used for calculating the needle tip was established.
After that, the software has been completely developed and integrated with the overall system software in during period two of the project.

Proof of concept of new biopsy system
Based on the identified requirements, the area in which the needle can move and the constraints and limitations imposed by the PEM imaging system and the breast immobilisation system, a general concept for the biopsy modules was developed. The strategy chosen to provide the necessary movement and positioning of the needle was based on a horizontal coordinated system also known as an Altazimuth system as shown on Figure 31. Needle movement description.

Control of all movements and communication and data handling with the other elements of the Mammocare system is provided by a computer controlled system based on the current Oncovision control system so simplifying integration of the biopsy system with the rest of the Mammocare systems.

To provide accurate position monitoring, both rotary stages were fitted with absolute ring encoders supplied by Renishaw, part No. RESA30USA115B

Figure 32. Location of encoder rings and reader heads.

Needle control movement as it is inserted and removed from the breast should be performed solely by the clinician with no powered movement of the linear axis. To accommodate this, the linear axis has been fitted with a hand wheel, to enable the clinician to manually insert/remove the needle, and a clutch to disengage the stepper motor to allow a smooth motion as the hand wheel is used.

Figure 33. Clutch system description

EVIVA biopsy device has a mounting foot incorporated into its body which allows the device to be located on a slide and locked in place with a plastic catch at the rear of the mounting foot

The control system is based on the current Oncovision control system so future integration of the complete system will be substantially quicker.

Figure 34. Biopsy electrical system design

The biopsy system has introduced a laser control system. The purpose of the laser is to measure the distance to the end of the needle point. We use this information also for do a needle calibration making a feedback of this information to the system.

The calibration protocol is based on the existing absolute encoder on the axis. The needle is rotated past the laser. At this point we would know if the needle is in its expected position.

If the needle is deemed to be in the correct position the laser or carriage would then be moved in order to find the tip of the needle. Once the tip has been found then the carriage will complete a oscillating function to verify it has reached the tip. We can also use the laser to check the needle width as well.

As we will have absolute sensors on both rotational axis and linear axis, these are used to cross reference the expected position of the needle.

Finally, all this hardware is commanded by a software system. This interface application is designed to perform the following tasks:

1. Home - Main Window (to Navigate)
2. Go to Home Position
3. Go to Relative & Absolute Positions
4. Get Positions
5. Get Status Report
6. Get Acquisition Report
7. Insertion – Start/Continue
8. Locate Laser Tip

Work Package 3. Biopsy Monitoring and Control Software Development.

The planning and navigation algorithms have been successfully developed meeting all the mechanical and clinical requirements. In case of conflict the clinical requirements have been prioritized trying to provide all needed information needed for the clinicians to perform the biopsy. The validation tests conducted show the reliability and consistency of the algorithms.

The software implemented has been developed to allow its compatibility with the GUI and the electronic control units.

From the functional point of view, the main conclusions are:
- The navigation software allows the user to reach a specific point inside the Field of View of the PEM system.
- The mechanical precision of the system has been evaluated and validated by the clinical experts at NKI-AVL.

After the tests, positioning validation and requirement validation, the successful integration with the user interface has proved the reliability and consistency of the algorithm

Monitoring and control software requirements
As a first step, the specifications of the monitoring and software requierement were defined, including:
• Complete process workflow
• Software requirement specifications (system overview, product functions, design and implementation constrains, user documentation, assumptions and dependencies).
• External interface requirements (user interfaces, hardware interfaces, communication interfaces).
• Functional requirements.
• Analysis diagrams
• Other non-functional requirements (performance requirements, safety requirements, security requirements, other requirements)

Graphical User’s Interface
The GUI has been designed to give the overall control to the user. This operation mode guarantees that the biopsy procedure is always performed under safety conditions and under supervision of the physician (who decides and selects the target volume to biopsy and when to perform any step of the procedure: start needle insertion, start tissue extraction or when to abort the biopsy process).

Oncovision has assured that requirements meet quality and usability criteria. NKI-AVL has provided support, from clinical perspective, providing usability criteria

Software is installed in two different workstations: an industrial embedded computer for acquisition and system control, and a GPU-based high performance station for reconstruction, which also holds the database. Both PCs are running on Windows operating systems: a Windows XP for acquisition PC and a 64-bit Windows 7 for reconstruction.

The components for MammoCare software are the following:

- MammoCare Database. This relational database hosts patient and acquisition details to coordinate the different software modules.
- MammoCare Patient Manager: This module allows creating, editing or deleting the patients from the system.
- MammoCare Acquirer: This module recovers the patient data from the database and performs the PEM acquisition and calibration.

Figure 35. MAMMOCARE Acquirer

- MammoCare Reconstructor: This module shows a 3D image from the acquired PEM data.

Figure 36. MAMMOCARE reconstructor

- MammoCare biopsy Viewer: This module loads reconstructed image and provides the interface for selecting the point where the clinician considers the biopsy should be performed. It also calculated optimum trajectory and gives the needle positioning coordinates for the motor controller module.

Figure 37. MAMMOCARE viewer

- MammoCare Motor control module: This module provides real-time monitoring of needle positioning and system sensors (paddle pressure, emergency, ring status). It provides positioning control of all mechanical subsystems: needle orientation and insertion depth, ring height and orientation, paddle position and pressure, ring aperture.

Figure 38. MAMMOCARE Motor module

Different GUI have been designed and developed, allowing the user to interact with the all the listed modules (except the MammoCare Database)

Software architecture and module dependencies are depicted in the following diagram:

Figure 39. Software Architecture

The resulting Graphical User Interface meets the main objective of providing overall control to the user, allowing the physician to interact with the different parts and functionalities of the system: PEM images acquisition, real-time 3D PEM visualization, trajectory planning for biopsy, real-time monitoring of biopsy procedure and control of biopsy positioning module.

The final implementation of the MAMMOCARE software has been divided in several independent modules, that address all functional and safety requirements. It has allowed the reduction of complexity and development and validation time. It has also added flexibility to the overall system.

Planning and navigation software
The resulting Graphical User Interface meets the main objective of providing overall control to the user, allowing the physician to interact with the different parts and functionalities of the system: PEM images acquisition, real-time 3D PEM visualization, trajectory planning for biopsy, real-time monitoring of biopsy procedure and control of biopsy positioning module.

The final implementation of the MAMMOCARE software has been divided in several independent modules, that address all functional and safety requirements. It has allowed the reduction of complexity and development and validation time. It has also added flexibility to the overall system.

Figure 40: PEM System Coordinates System
Figure 41: MammoCare System Coordinates system

The main part of the algorithm will be implemented in the Motor Manager module. This part will have as input the MammoCare Coordinates which are the output of the Biopsy Viewer Module. The Needle path calculation will calculate the next position.

To calculate the next position to move, the algorithm need several mechanical inputs, as the azimuth offset, the elevation offset and needle length, needle type...

To validate the needle position we have designed a tool that allows to place a radioactive source with 1/10 of millimeter error. To perform the position validation 25 positioning have been performed.

They have been performed on different positions been the most studied the ones that were on the FOV limits. Less than 0.6 mm error was obtained with a 4mm thick needle and a Biopsy hole of 12mm.

Figure 42: Radioactive source holder to verify needle position

From the functional point of view, the main conclusions are:
- The navigation software allows the user to reach a specific point inside the Field of View of the PEM system.
- The mechanical precision of the system was evaluated and validated by the clinical experts at NKI-AVL.

Work Package 4. Systems Integration.

In this work package, the PEM imaging and biopsy guiding modules have been integrated into one operative PEM-guided biopsy system. The integrated MAMMOCARE system has been calibrated and verified through in-vitro validation of the overall system’s performance.

Assembly and set-up of different subsystems
The main structure that was designed to give support to all subsystems was based on a rotation plate. This Rotation plate was designed for support the PET detector, the PET open and close mechanisms, the compression paddles and the biopsy subsystem

The compression paddles were integrated, including load cell cabling, safety clutch, knob mechanisms and final paddles system.

Figure 43: Compression paddles integration

The PET open and close mechanism was installed on the main rotation pallet including half rings support, stepper engine, driver, end line switches and safety protection.

PET detector integration. The original proposal was to install a single ring, but in order to provide a better performance, the development team decided to build a double ring detector. The development team also improved the original detector external design.

Figure 44: PET detector integration

Biopsy Robotics integration. The final biopsy robotics setup received was bigger and heavier than first designs.

Calibration Laser. The calibration needle laser position was changed for avoid future problems with the robotic arm movements.

During the integration and test procedure it was installed using an adjustable support. This support was changed to a fixed one and a cover was added for protect the laser detector to external lights.

Computer and electrical integration of all subsystems. All subsystems are supported by a main electrical box, where the electrical panel, the medical grade isolation transformer, the filters and the electromagnetic disruptor are installed.

Figure 45: Computer and main electrical panel integration

Global system’s calibration
MAMMOCARE system full validation needs to be done on 3 steps. These 3 steps are based on separate validations of each subsystem, the PET subsystem as our imaging system, the biopsy subsystem as a robot system and finally a coordinated fusion between them.

PET system calibration and validation was done during T1.5 Proof of concept of new biopsy system and described on D1.3 Proof of concept of PEM imaging system compatible with biopsy, obtaining as conclusions that the new detector developed for MAMMOCARE prototype preserves performance of MAMMI product design and provides an open design to perform biopsy.

The detector subsystem has being characterized on the 2 active configurations: open and closed ring.

We need to check the cross correlation between both detector actives modes. Final cross check between both detection systems must verify that coordinates obtained are the same.

Finally we need to perform a needle calibration. This operative is done each time Motor Software starts the initialization of the system, correcting deviations and preventing malfunctioning situations. If the distance is acceptable, the system will use this value as new calibrated position.

In-vitro system's validation
The final step for a full validation of MAMMOCARE system is to ensure that the coordinates provided by the optical system on open ring mode have a good cross correlation with the coordinates given with mechatronic system.

To validate the needle position the development team has designed a tool that allows placing a 0.5x0.5x0.5mm radioactive source inside the FOV. The user can move this radioactive source on X an Y axis with 1/10 of millimetre error.

Figure 41: MammoCare System Coordinates system

Re-design of components and connections
During first test, the development team had to make different redesign tasks for functionality or safety reasons:

The final prototype rotation pallet weight (~100kg) plus his final length, was unexpected from first designs. This 2 main issues caused a bigger torque, more inertia and bigger friction taking our system on the limit of operative range

A redesign of several components in order to take the system again in operative range was needed

Up and Down: The development team improved the up and down movement optimizing the elevation actuator and balancing the total weight with 4 springs
Turning movement: Inertia increased and a change of rotation stepper motor into an AC rotation plate was needed

During the first assay, different safety circuits were tested. The compression paddles were designed with a detector that informs the software that the paddles are not completely out and must not move. Due the high risk of damage to the patient during biopsy procedure in case of a failure of this first system, a secondary system was designed.

This secondary circuit avoids any elevation column movements if compression paddles are not completely out.

Regarding the high inertia on the rotation plate and in order to avoid any undesired rotation movement, also during a black out, a brake system was installed in order to add an extra safety system.

Work Package 5. Validation of MAMMOCARE System.

In this work package, the performance of integrated PEM-guided biopsy system was validated in order to evaluate its feasibility for clinical application in patients. For this purpose, NKI-AVL, as end-user, conducted a clinical study with several patients.

Protocol definition.
All patients will be informed of the aims of the study, the possible adverse events, the procedures and the mechanism of treatment allocation. Patients will be informed by a physician and a Nurse Practitioner and will be given sufficient time to decide on participation. The patient has the possibility to contact with an independent medical doctor.

Documented written informed consent will obtained before patients are included in the study. This written informed consent will be kept in the patients medical records. The patients must be assured that their identity will be protected

The primary objective of the present study is to assess the feasibility and safety of a newly developed PET-guided breast biopsy system in breast cancer patients. Other secondary objectives:

- Obtaining information about the duration (in minutes) of the 18F-FDG-guided biopsy procedure.
- Comparison of the gene-expression and molecular subtyping between the breast biopsies taken from the tumour region with the highest 18F-FDG uptake and the tumour border.
- Measuring the radiation exposure of the physicians during the biopsy procedure.

Protocol
After the conventional PET/CT scans, patients with demonstrated 18F -FDG-avid breast lesions will be transferred from the Nuclear Medicine department to the MAMMOCARE intervention room at the sixth floor (room number 6B.21). In here, 18F-PET-guided biopsies will performed by the Radiologist and Nuclear physician based on the following protocol (Figure 1):

1. PET image acquisition of uncompressed full breast
2. Calculation of best plane for biopsy approach: Physician marks the centre of the volume of interest on the visualisation screen of the Graphical User’s Interface (GUI) and the planning calculates the best plane to slightly compress the breast (perpendicular to biopsy approach).
3. Immobilization of the breast: Compression plates positioning system will rotate the plates to the selected plane and closes them to compress the breast until reaching the optimal pressure threshold for breast fixation. Biopsy approach will be performed perpendicular to the calculated plane for the plates. Plates will have a window in the middle to allow the biopsy needle access to the breast. Prior to biopsy needle insertion 10 mL of 1% lidocaine will be administered as local anaesthesia.
4. Planning of needle trajectory: Once the breast is fixated in the proper position, the system will calculate the best trajectory for the needle from the skin to the target lesion. Physician will mark the centre of the target lesion on the visualisation screen of the GUI and the planning software calculates the best entry point and trajectory for the needle.
5. The system will obtain the initial spatial 3D coordinates of the lesion and the needle tip, both in the biopsy reference frame. Calculations for the optimum path (lower distance from the outside and with lower risk) will be made and the offset values for each degree of freedom (DOF) of the biopsy positioning system (Ry1, Dy, Rx, Ry2, Dz) will be displayed in the GUI.
6. Motion of the needle to the entry position: The physician will adjust (or will order the adjustment automatically) the position so that the offset values are zero in all DOFs, except for Dz (the longitudinal displacement of the biopsy needle).

The position and orientation of the needle will be visualised in real-time on the monitoring screen (navigation software), so the physician can continuously check the needle positioning process. At this point, the needle will be aligned with the planned insertion path and the needle tip will be positioned at the entry point outside the breast.
7. Needle insertion: the physician will start the insertion process (Dz) through the GUI command guiding the displacement of the needle through the breast tissue, towards the target point. At this stage, the rest of DOFs are restricted. During the insertion process, the real trajectory of the needle will be continuously shown on the visualisation screen by means of an augmented reality visualisation tool. Eventually extra lidocaine 1% solution could be administered during needle insertion if necessary.

8. Monitoring and control of needle insertion: Apart from allowing real-time visualisation of insertion process, navigation software will continuously enable to check for misalignment of the needle. In the case of needle deviation above a “safety” threshold, then the navigation software will send a warning message. When this happens, identification of target position and subsequent trajectory planning should be repeated (physician should mark again the target point on the screen and the planning software will modify the trajectory in order to reach the lesion position). This insertion control step will be repeated in a close loop until the needle tip reaches the target point for biopsy.

Subsequently, the physician will control if the needle is placed in the target point before ordering biopsy. The physician will decide the start of biopsy (tissue extraction) from the GUI related command. Three biopsies, which is standard clinical practise at the NKI-AVL, of the primary breast tumour are performed. Two biopsies are taken from the area with the highest 18F-FDG uptake and one biopsy is taken from the tumour border using a commercial 9-gauge vacuum-assisted needle (Eviva, Hologic, IN, USA).
9. As safety check, a hydromarker will be inserted to indicate the biopsy location. Mammograms can visualise if the biopsy location (hydromarker) correspond to the location of the tumour. Furthermore, the hydromarker is needed as orientation point to insert the 125I-marker.

10. Finally, a dose calibrator will be used to measure the presence and quantity of 18F-FDG in the breast biopsy specimens.

System installation at end-user facilities
Installation was started at 8th of June at NKI facilities. Installation process plus calibration and validation of close an open ring take to the development team 2 weeks.

Once this first part of installation procedure was finished, training sessions to NKI personnel and radiologist started in order to keep final feedback and on site software corrections or modifications.

System calibration and validation
In vitro validation protocol includes as first step the analysis of the reproducibility and accuracy of open and closed ring image acquisitions. Final conclusion was that the MammoCare biopsy prototype produces reproducible images, so changes in breast lesion coordinates during the biopsy procedure can in fact be assigned to an actual movement of the lesion.
Errors between the image coordinates of open and closed ring acquisitions were smaller than the 2 mm voxel size of open ring reconstructions. These errors are small enough to locate the breast lesions using the best possible precision in the Biopsyviewer software (1 voxel)

Compression paddles system was validated, and also the readout of the loadcell, which is connected to one of the compression paddles and is visualized in the MotorModule software. Final conclusion was that the Mammocare system provides an accurate compression force measurement, especially in the 70-120 Newton range. The 70-120 Newton range is the expected working range that will be used to compress breasts during the biopsy procedure

The last validation test was to run through all 5 steps of the clinical protocol and measure the accuracy of the biopsy needle positioning system.

Figure 46: Photographs taken from the Eviva biopsy needle without (A) and with (B) the piece of plastic inside the opening of the needle

The final absolute error, taking into account the error in all directions sqrt(x^2+y^2+z^2), in the worst case is under 3 mm.

Clinical validation study
The Mammocare system was ready for clinical validation after evaluating all steps of the clinical protocol in a phantom setup. An interesting group of patients was selected for evaluating the dedicated PET breast images obtained with the Mammocare prototype.

In total, 8 patients who were scheduled for neo-adjuvant chemotherapy were scanned on the Mammocare prototype. Tumour visualization on Mammocare was appropriate for eventual biopsy in 5 out of 8 patients. In one patient the lesion was not optimally displayed due to its location in the vicinity of the chest wall and in two patients FDG-uptake was too low for precise tumour visualization (conventional PET/CT was also not able to visualize the breast tumours in these 2 cases).

Three biopsy specimens are obtained per patient as standard clinical protocol in the NKI-AVL. Ideally, those three biopsies are taken from the three intratumour regions with the highest 18F-FDG uptake. Therefore, the following data was extracted from the 5 out of 8 Mammocare patients;

1) Size of the tumour on MRI: the size of the breast tumour is often much larger than the size of the intratumour regions with the highest FDG-uptake.
2) Distance between tumour and chest wall: it is often difficult to detected breast cancer which are located near the chest wall
3) Intensity of breast tumour on Mammocare images: this is visually scored as none, mild, moderate, or high.
4) Heterogeneity of 18F-FDG uptake in tumour: this is visually scored as none, mild, moderate, or high.
5) The size and coordinates of the intratumoural regions: data up to three intratumour regions are given
6) Distances between the intratumoural regions with the highest 18F-FDG uptake

Overall the intratumoural regions having the highest 18F-FDG uptake were much smaller than the size of the whole breast tumour. The Mammocare images showed at least three regions having a higher 18F-FDG uptake in 4 four patients, and 2 regions in one patient. Breast compression forces between 70-90 Newton were considered acceptable based on patients feedback. Finallty the Mammocare system is able to visualize small (<10 mm) intratumoural regions with a high 18F-FDG uptake, which are considered ideal as biopsy target in breast cancer patients who are scheduled for neo-adjuvant chemotherapy.

Final conclusions
The MammoCare biopsy prototype provides;
- Reproducible images
- Errors between open and closed detector ring acquisitions that are smaller than the image voxel size.
- Accurate breast compression measurements
- Detailed information of small intratumoural breast cancer regions which are having a higher 18F-FDG uptake can be selected for biopsy by the Mammocare system.

Work Package 6. Innovation and Related Activities.

This workpackage deals with all aspects related to dissemination, exploitation and protection of the results of the project.

The Consortium created a first version of the Plan for the Use and Dissemination of the Knowledge during the first reporting period, delivered to the REA as deliverable D6.2. This deliverable describes the future planned actions to be done by the Consortium related to the use and dissemination of the project results. It also includes a list of actions already performed and an analysis and specifications of the formative/training needs of each SME, depending on the result.

During the second reporting period, the consortium created and submitted to the REA, as deliverable D6.3 a final version of the Plan for the Use and Dissemination of the Knowledge.

The main milestones achieved during the project are:

• Creation of a website of the project
• Creation of all the dissemination material: brochures, press release templates, presentation, etc.
• Press release publication (D6.1): the text of the press release was produced and agreed by all the partners. It provided relevant information about the project objectives and its participants. It was published in the partner’s websites. It had a major impact in Spanish media.
• Publication in the partner’s website of information about the meetings that have taken place during the whole duration of the project
• Appearances in media:
o GEM IMAGING appeared in the news of Canal 9 (regional TV of Valencia) on October 17th, speaking about the MAMMOCARE project
o The News Programme of the TV of the Polytechnic University of Valencia made a report about the MAMMOCARE project. IBV represented the MAMMOCARE consortium
• HERI published a press release in a local newspaper
• The consortium created and published a promotional video of the project: https://www.youtube.com/watch?v=vk2gnHv1sak
• The consortium issued a press release regarding the end of the project, with a significant impact in Spanish media.

The MAMMOCARE consortium has identified several dissemination activities to be carried out in the short and mid-term, after the end of the project, including the participation in conferences/exhibitions, publications and web/social media activities.

Work Package 7. Project Management.

The management of the project includes, among others, the following activities:

• Communication and reporting to the REA
• Communication among the consortium
• Monitoring of deliverables, milestones and reports
• Preparation, update and management of the consortium agreement
• Ethical issues management

At the beginning of the project, the project Coordinator was changed. Initially, Dr. Luis Caballero assumed this role, but he left Oncovision one month after
the start of the project. The new project Coordinator is Mr. Jorge Álamo. This change was communicated to the REA project officer. An amendment to the
Grant Agreement was required, that included also a change in the contact address for Oncovision.

The communication between the partners has been very intense during the project. Four meetings have been organized and several online conferences:

Kick-off meeting. Valencia (Spain). 4th of October 2013
Technical progress meeting 1. Valencia (Spain). 11th of March 2014
Technical meeting (Teleconference using Webex). 5th of February 2014.
Technical meeting. (Teleconference using Webex). 18th of November 2013. Technical meeting. Valencia (Spain). 19th of November 2013
Technical meeting. (Teleconference using Webex). 27th of January 2014
Technical meeting. Valencia (Spain). 11th of April 2014
Technical meeting. Melton Mowbray (UK). 26th of September 2014.
Technical meeting. Amsterdam (Netherlands). 26th of June 2015.
Technical meeting (Teleconference using Webex). 22nd of July 2015.

Additionally to the Consortium meetings, bilateral meetings have been organized for specific technical topics:

Technical meeting: GEM IMAGING and HERI, on November 18th 2013, through Webex
Technical meeting: GEM IMAGING, CSIC and IBV, on November 19th 2013, at IBV’s premises
Technical meeting: GEM IMAGING and STATICE, on January 27th 2014, through Webex
Technical meeting: GEM IMAGING, CSIC and IBV, on April 11th 2014, at IBV’s premises
Technical meeting: GEM IMAGING and IBV. August 3rd 2014, at GEM IMAGIN’s premises
Technical meeting: GEM IMAGING and IBV. 10th of March 10th 2015 at IBV’s premises
Technical meeting: NKI-AVL and GEM IMAGING. June 8th 2015 at NKI-AVL’s premises.

One of the main objectives of the management is to guarantee the achievement of project objectives through technical and economic management and
coordination of different activities within the Consortium. In this sense, the MAMMOCARE project has achieved the objectives for the project
and it’s in line with the workplan as approved by the REA.

A web page has been created to explain the project aims and objectives and to disseminate information about project activities and results. As a dissemination
vehicle, the website includes news about the project. MAMMOCARE website also make their deliverables available. MAMMOCARE website has been designed to be attractive and easy to use, with intuitive navigation. It will be up to date and submitted to key searching engines.

The website address is: http://www.mammocare.eu.

Potential Impact:
The MAMMOCARE project was designed as a first approach like a Breast biopsy system guided by Positron Emission Mammography allowing real-time 3D visualization of tumor lesion and needle insertion guidance for higher sampling accuracy and efficiency.

The key point on this project is to provide a prototype ready for adaption to the production line that can provide an answer to the following key points:

- Actual biopsy procedures are based on ultrasound systems or (the most advanced systems) on X-Ray visualization systems. It is demonstrated that PET visualization systems provide the lower rate of false positives. The best biopsy systems should be based on PET technology
- Actual PET systems have lower spatial resolution than stereotactic systems (X-Ray). In order to have the best possible spatial resolution we will need to use the dedicated breast PET with better spatial resolution
- During biopsy procedure, it is not possible to check if it is being done on the selected zone. The best biopsy system must have the capability do real time imaging and correction trajectory in order to ensure correct procedure
- The best biopsy system must be comfortable for the patient
- The best biopsy system must reduce to minimum the injury during the procedure
- The best biopsy system must me ergonomic for the radiologist; software and mechatronics must contribute to an optimized procedure and precision.

Breast cancer is a major public health problem. Development of technologies for earlier cancer detection and more effective treatment of the disease provide clear benefits for European patients and Healthcare Systems: Reduction of mortality risk, Improvement of patient’s quality of life, Costs savings for the Healthcare systems.

Aiming to go in line with that philosophy shows specific benefits of using MAMMOCARE system will be:
1- More than 30.000 women will benefit (by 5th year) from the exclusive ability to detect and biopsy cancerous lesions that are not visible by conventional imaging techniques, allowing an earlier cancer diagnosis.
2- More accurate tumour sampling thanks to real-time monitoring and control of needle motion as well as the ability for correction of needle trajectory. This avoids the need for a second biopsy process related to inadequate sampling and positioning errors. Considering that 8% of biopsies require a second trial due to positioning errors and a total number of 1,3 million biopsies performed
each year in Europe, 100.000 repeated biopsies could be avoided by using MAMMOCARE.
3- Optimal tumor sampling and subsequent treatment definition, thanks to the PEM capability to differentiate most active metabolic regions within the tumor. Treatments optimization will also reduce the associated costs, like reduction of the total number of radiotherapy or chemotherapy sessions.
4- Reduce the duration of biopsy procedure, thanks to real-time imaging and monitoring of the process.

The aim of MAMMOCARE system is to reduce at least a 50% the duration of the whole biopsy process with regard to existing competing systems (which takes more than 1 hour according to the experience of consulted clinicians). This time saving is also equivalent to costs saving for the hospitals and National Healthcare Systems. Considering a total amount of 1,3 million biopsies performed yearly in Europe, 30 minutes of time reduction per procedure yields a total time reduction of 27.083 complete days (74 years!). Considering an average gross salary for the biopsy physician of about 35.000 €, this would mean a salaries saving around 2,6 M€ per year.

Because of the threatening growing trend in breast cancer, the consequence of inefficiencies in current image-guided biopsy equipment for diagnosis of breast cancer will result in significant economical impact for the European Public Health systems, but also offers an opportunity for the SMEs participating in the MAMMOCARE project to better meet the end-users needs.

Dissemination strategy.

The consortium has already identified key target groups for dissemination. The EDC is working to ensure that the Dissemination and Awareness campaign of the project reaches the five following groups:

1. The General Public through patients organizations.
2. The International Scientific community as for example the European Conference on Nuclear medicine and the European conference in breast cancer.
3. The Clinicians, through different clinical networks of the clinical partner of the project.
4. The Health and Education Professionals.
5. The industrial stakeholders working in the filed of medical imaging and biomarker development.

Dissemination activities
The MAMMOCARE consortium has identified the following communication channels for the dissemination of the results of the project:

- Publications: All the partners are encouraged to publish scientific papers on the most outstanding technical and scientifical results. The RTD performers and the end user (NKI-AVL) will be responsible of the writing of these scientific papers. They are very keen on this dissemination via as the number and quality of their published scientific papers is an important performance indicator for them.
- Articles for popular media. These articles aim to promote the project results to a broader public. The Dissemination Manager will write these articles, but each partner will be responsible of translating them to their language and disseminating them via their local media.
- Participation in conferences, workshops and trade fairs: The MAMMOCARE partners are encouraged to attend to the most relevant conferences, congresses, workshops and trade fairs to present the results of the project. The EDC has prepared some of the materials that will be distributed to all partners for such purpose (brochure, presentation). All conferences, talks, posters, etc.… will have the approval of the EDC. Some potential events are: Annual meeting of the SNMMI (Society of Nuclear Medicine and Molecular Imaging), Annual Congress of the EANM (European Association of Nuclear Medicine), MICCAI (Medical Image Computing and Computer Assisted Intervention Society), IEEE NSS/MIC, SPIE (international society for optics and photonics) Medical Imaging, Congreso Nacional de la Sociedad Española de Senología y Patología Mamaria.
- Radio and TV. The Consortium partners are encouraged to participate in local, regional or national radio and/or TV media to disseminate information about the project.
- Contacts with other projects.

The MAMMOCARE consortium has identified the following dissemination activities where the partners could participate in the short and mid-term:

- Participation in conferences/exhibitions:
o EANM'15 - 28th Annual Congress of the European Association of Nuclear Medicine. October 10 - 14, 2015. Hamburg, Germany.
o 2015 IEEE Medical Imaging and Nuclear Science Symposium (IEEE NSS/MIC, http://www.nss-mic.org/2015/public/welcome.asp). October 31 - November 7, 2015 in San Diego, USA.
o 2016 IEEE Medical Imaging and Nuclear Science Symposium (IEEE NSS/MIC). October 29 - November 6, 2016 in Strasbourg, France.
O Annual meeting of the Radiological Society of North America, RSNA 20145 November 29 to December 4, 2015 in Chicago, USA. http://www.rsna.org/Annual_Meeting.aspx
O San Antonio Breast Cancer simposium. December 9th to 13th in San Antonio, USA. http://www.sabcs.org/
O SPIE Conference. February 26 to March 3, 2016 in San Diego, USA. http://spie.org/medical-imaging.xml

- Publications: article in "Revista de Biomecánica" of IBV, at the end of the project.

- Web/social media: all partners are encouraged to distribute information about the Project results through their Internet channels:
• Website
• Twitter
• Facebook
• Linkedin
• Google+
• Youtube

Results of the project
All the knowledge generated in the project will be transferred by the RTD performers to the SMEs. Therefore, these will be the organizations that will make a commercial use of the results of the project.

There are three results coming out of this project, that will be integrated in the MAMMOCARE system:

• R1: 3D real time PEM imaging system (patent), with a mechanical design compatible with the new biopsy system, as well as foreground related to any further developments in technical specifications of the device as may result from engineering design of the coupling of the biopsy system with the imaging detector, will be owned by GEM IMAGING. The MAMMOCARE system will be patented by GEM IMAGING, who will be the integrator and distributor, and will give preferential rights to STATICE SAS for the manufacturing and supplying of the biopsy system (R2) and to VA for supplying the biopsy monitoring and control software (R3), as long as they meet appropriate quality standards, and according to the agreements to be reached between parties. Any foreground related to testing and validation of the imaging system according to regulatory standards that may result from testing and validation activities will be also own by GEM IMAGING.

• R2: Biopsy positioning module (industrial design), compatible with the PEM visualization module, will be able to communicate with the control software, in order to assure the precision mechanism and the dedicated breast compression frame will be oriented in the optimal path, with the right level of breast compression. The precision mechanism will be fully compatible with commercial current vacuum-assisted breast biopsy instrumentation. This result will be share-owned by STATICE SAS and GEM IMAGING. STATICE SAS will exclusively supply the biopsy system to GEM IMAGING for its integration in the MAMMOCARE system, according to the agreements to be reached between parties. STATICE SAS will also have exploitation rights for other non-competing markets, having explored image-guided biopsy of liver, lung and thyroid fields as a secondary market

• R3: Biopsy monitoring and control software (copyright), will be the part of the system’s software dealing with planning, monitoring and control of the biopsy procedure. This includes the GUI to give the overall control to the physician and allow the follow-up of biopsy procedure; as well as the planning and navigation software packages aimed to calculation of optimal needle insertion path and real-time monitoring of needle motion with regard to the target lesion. This result will interact with R2 for the tracking and control of needle motion and with R1 for the real-time calculation of 3D position of the target lesion. This result will be share-owned by VA and GEM IMAGING. VA will license its use to GEM IMAGING for its integration in MAMMOCARE system. VA will also have exploitation rights for other markets, except for breast cancer PEM-guided biopsy.

Exploitation plans
The result of the MAMMOCARE Project is an integrated breast biopsy system guided by PEM Imaging. The consortium has developed a prototype that has been installed at NKI-AVL premises to conduct several clinical validation tests. Therefore, the result of the Project has a Technology Readiness Level (TRL) 7: System prototype demonstration in an operational environment.

In order to start the commercialization of the final product, the partners (the SMEs) still have some futher steps to carry out:

1. Optimization of the system: this includes the inclusion of the latest photo-detector technolgy and the corresponding adaptation of the electronics of the system and improvement of the reconstruction algorithms, as well as the optimization of the system ergonomy
2. Improvement of the system final design (external look of the system) to make it more attractive
3. Detailed commercialization plan, including the definition of the sales strategy (geographical segmentation), allicances policy (direct sales, distributors…), price policy, among other aspects.
4. Clinical validation of the system. This is very important to allow the commercialization in some countries, like Germany or the USA
5. Obtention of the CE and FDA approvals, that allow selling medical devices in Europe and the USA.
6. Industrial scale, to define all the internal processes to manufacture the equipment

To achieve all these objectives, the Consortium is planning to submit a proposal to the next call of the Fast Track to Innovation instrument, under H2020.

Although the partners have to prepare a detailed commercialization plan, there are some important market objectives already identified:

- Germany: this is the most important market in Europe.
- USA market: this is the biggest market in the world.

The partners will focus their efforts to obtain clinical evidence in these countries. This is extremely important to sell medical equipment in them. As a next step in the overall strategy, the partners will approach reference centers in those two countries to run clinical trials with MAMMOCARE.

- GEM IMAGING

For GEM IMAGING, this project will allow extending its product’s portfolio: the launch in the market of an innovative product to solve a huge health problem to the users will therefore increase the perceived value of the company in the final market. The company has a commercial PET line of products (MAMMI, ALBIRA). GEM IMAGING has as a clear strategy offering organ specific PET products. The MAMMOCARE system will be part of this line.

- STATICE

For STATICE SAS, this project will allow extending their applications to the medical sector (product differentiation), which will increase not only their annual turnover, but will also allow the alliance with companies addressed to large markets (medical devices and diagnostic imaging).

The acquired know how in micro displacement and mechanical control efforts in a complex environment will allow strengthening STATICE’s medical activity and development of its mechatronics department. Therefore, STATICE SAS will also benefit by offering an accurate biopsy system for other non-competing markets (MIS for orthopaedic surgery, image-guided biopsy of liver, lung and thyroid), as a secondary target market.

- Virtual Angle

Virtual Angle BV aims to extend its current product portfolio among the Medical and Clinical healthcare products offered, giving a strong step forward in real-time navigation/monitoring software and towards a stronger market positioning. The outcomes of MAMMOCARE will allow VA to adapt their developments to other clinical applications that need biopsy guiding monitoring or other types of surgery that may benefit from the use of navigation software.

Virtual Angle B.V. offers Cloud Computing services and solutions, via the Virtual Angle platform, which is primarily based in company lines of software products. One of the available solutions is based in the in-house developed Virtual Clinic platform (http://vclinic.virtualangle.com/) which will be expanded and diversified in order to better integrated the outcomes of MAMMOCARE.

Within the scope of the MAMMOCARE project VA is also aiming at exploiting the MAMMOCARE monitoring software in the space sector - VA is a European Space Agency supplier with strong involvement in ESA medical studies for manned flights and long term isolation.

VA also intends to explore the outcomes of the MAMMOCARE project within the veterinary sector combined with a new software platform that will be released during the upcoming year.

Due to the company involvement in the civil protection and disaster management activities internal preliminary discussions are taking place, at management and R&D company levels, regarding the possibility of exploiting the project results (namely the monitoring software) in the shape of a rapid deployment platform to be applied in surgical applications in collaboration with other partners and/or internal developed tools.

The commercial dissemination related with the MAMMOCARE outcomes will be initiated in the second year of the project and it will involve collaboration with the VA partners and distributor network.

Aside of the commercial independent activities, VA will also collaborate with GEM IMAGING and Statice in a joint market approach effort.

List of Websites:
www.mammocare.eu

Final Report Summary - MAMMOCARE (Breast biopsy system guided by Positron Emission Mammography allowing real-time 3D visualization of tumour lesion and needle insertion guidance for higher sampling accuracy and efficiency)

Related documents

Share this page

Download