#	Pagina
attuale pagina	/open-h2020/projects/199861/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - MURAB (MRI and Ultrasound Robotic Assisted Biopsy)

Teaser

Summary

The MURAB project deals with the diagnostic biopsies, focusing on breast and muscle application. The main issues that MURAB addresses is the complicate setup and execution of a magnetic resonance imaging (MRI) guided biopsy and the lack of accuracy in ultrasound (US) guided biopsies, especially in small lesions. Sampling error in oncology is an important issue. Questions that are often unanswered are: How sure are we that we hit a certain target? Is the imaging reliable enough for guidance? Did we hit the most aggressive part of a certain lesion?
Aimed at both muscle and breast research, but possibly suitable for other applications, the MURAB project combines the high sensitivity of MRI with the high accuracy and availability of US guided biopsies. It is anticipated that image fusion and robotic needle guidance will increase the detection rate of target US and the need for costly MRI-guided biopsies will decrease. The quality of patient care will improve because the detection rate on target US will improve. The clinical workflow / planning of patients for target US is more flexible compared to MRI guided procedure, not to mention the lower cost and pain level, making this the main advantage of MURAB.
A new workflow will be offered to the practice with reduced use of expensive MRI and at the same time yield the same precision during samples targeting. Guided by a novel MRI- US registration, a robotically steered US transducer equipped with an acoustically transparent force sensing will autonomously scan the target area and optimally acquire volumetric and elastographic data. Additionally, parts of the system functionality can also be useful in other clinical workflows.

Work performed

University of Twente coordinated the management activities needed to run a successful and productive project. Weekly telephone conferences were organized to exchange information about progress, unforeseen challenges and to make decisions for the next steps. Several project meetings took place in different locations and in concomitance with workshops, conferences and summer school, as described in the technical report.
The MURAB project was presented at the KUKA Corporate Research Open Days, Bessensap 2016 event, InScience Dutch International Film Festival 2016 and to the following conferences: European Conference of Radiology 2016 and 2017, Technical Innovation in Medicine Conference 2017, European Society of Breast Imaging 2017, International Ultrasonics Symposium 2017, ERF 2016 and 2017, BioMed 2016 2017, CRAS 2016, 2017, 2018 , CARS 2016, IROS 2017, at the COSUR 2018 summer school in Verona MURAB had exhibition stands at IROS 2018 in Madrid and at the MEDICA Trade Fair in Dusseldorf, where KUKA presented the MURAB robotic setup acting during an US acquisition on a realistic breast phantom. During the Nemo Science Live event that took place from 20 to 26 August 2018 at the Nemo Museum in Amsterdam, US technology included in the MURAB system was tested on more than 100 volunteers with 3D US of the front lower leg muscle.
Realistic breast phantoms were designed to investigate the MRI interfacing for the proper acquisition of the data, as well as the US imaging. MRI and 3D US measurements for both breast and muscle on healthy volunteers was performed. A cone structure was designed to scan the breast with a Siemens ABVS transducer. Another acquisition method makes use of the KUKA robotic arm by moving the US transducer to get the B-mode scans. The 3D US volume is updated after each B-mode image acquisition. Each image was integrated into a target volume using a modify nearest neighbor algorithm. The algorithm was tested in RUMC as described in the technical report.
A cross-correlation-based tissue motion estimation algorithm was tested to estimate the lesion displacement induced by needle insertion or other external motion.
US images and MRI images were registered using the external surface of the tissue on both breast and muscle. The geometric registration combined affine and thin plate spline functions, while a second approach based on finite element method (FEM) deformation was also tested. Due to the real-time constraints that are hard to be accomplished with the FEM modeling, a geometry-based dynamics (PBD) algorithm was also implemented.
Multimodal markers (i.e. visible in different diagnostic images such as MRI and US) were designed. They are fully compatible with the robotic scanning and are used to support the scanning trajectory alignment and the image registration process.
Several prototypes of the end effector were designed. The end effector was integrated with the KUKA robotic arm.
The robotic head is equipped with stereo camera and a small light projector that helps in the localization of the multi-modality markers attached to the skin. A 3DOF mechanism was implemented to aim the biopsy needle in the plane of the US probe which is rigidly attached to the robotic arm.
An implementation of an automatic trajectory to be followed by the robotic arm equipped with the end effector was tested. Hybrid force-motion control for reactive control needed during the scanning phase was designed.
An innovative sensor that enables a 2D linear US probe to do elastography measurements was designed. The sensor consists of an acoustically transparent material that envelopes the US probe and the pressure at the tissue-US probe interface is determined by analyzing the US image.
Each of the components of the MURAB system was tested in pre-clinical (in vitro) environment. An in-vivo test that quantified the amount of deformation during subsequent MRI scan of the same patient was performed.
The clinical requirements were defined f

Final results

Several progresses beyond the state of the art were achieved: new methods for 3D US acquisition and reconstruction, real-time methods for multimodal image registration, new methods for the calibration of the deformable models and compensation of the deformations, prototypes of a robotic head for breast and muscle biopsies, new methods for the control of a robotic arm to perform the 3D US scanning of a soft tissue, identification of materials for the realization of multimodal markers, a new pad sensor to be attached to the US probe for elastography measurements.
Some of these realizations may become commercial products by themselves (e.g. multimodal markers, image acquisition system, US pressure sensor).
Most of the technologies developed so far were integrated in the MURAB robotic system and tested on realistic phantoms of the breast and of the leg. The consortium is organizing the hardware and software for clinical trials and the for the approval required by the ethical committee. A complete analysis of the market and of the business model was performed by Siemens. This analysis comprised studies about: potential markets, market size, costs of the MURAB procedure and comparison with the cost of a standard procedure, number of MRI guided biopsies in the countries involved in the consortium and the potential number of MURAB installations within a country.
Also, the possibility to extend the MURAB system to other cases (e.g. liver, kidneys) may lead to a major socio-economic impact of the project.