Interdisciplinary Project Abstracts

Multiscale model based on medical imaging and influence of the KLF10 gene on the biomechanical behavior of skeletal muscle

Striated skeletal muscle tissue is a complex, hierarchical structure composed of a large number of multiscale components influencing the behavior and overall mechanical response of the organ. Majority of studies focused on macro- or microscopic scale as the recent research studies conducted at the Mayo Clinic and the UTC have highlighted the influence of the transcription factor Klf10 (Krüppel-Like Factor 10) on the passive and active mechanical properties of muscle fibers. However, its implication on the global mechanics of the entire muscle remains totally unknown. The MYOMETISME project target aims at obtaining ex vivo and in vivo experimental data at different scales for the modeling of the skeletal muscle.

• Co-PIs: S. Chatelin & S. Bensamoun (ICube & UTC)

CRYO-Track: Guiding multiple needles on planned trajectories

The aim of the CRYO-Track project was to launch a collaboration between ICube's IMAGeS research team (Caroline Essert) and the Medical and Environment Computing team (MEC) of the Technischen Universität Darmstadt (Anirban Mukhopadhyay) on the subject of computer assistance for percutaneous procedures. The objective of the project was to study needle insertion and real-time guidance during percutaneous cryoablation in abdominal structures in order to improve both the safety and the efficiency of a procedure that is currently performed without direct visibility.

• Co-PIs: C. Essert & A. Mukhopadhyay, ICube & TU Darmstadt

Prevention of aging and early rupture of breast implants

Breast implants are made of a silicone shell which is a polydimethylsiloxane elastomer. Breast implants, like all implantable medical devices, are subject to wear and tear as they age in the body. In addition to this wear and tear, they are subject to external trauma, whether accidental during daily life, during insertion or during mammography. The kinetics and mechanisms of aging are poorly studied, and the phenomenon of implant rupture is not elucidated. The PRIMA project is about the mechanical characterization of breast implants and their evolution over time includes not only a mechanical modeling of rupture, but also experimental mechanical characterization of both the membrane and the gel constituting the prostheses.

• co-PIs: N. Bahlouli, S. Chatelin, F. Bodin & E. Breton, ICube   

Parsimonious technologies and methods for upper limb rehabilitation

In recent years, the development of robot-assisted therapies for the rehabilitation of neuro-motor function disorders has opened the way to novel technologies to assist the patients and the practitioners during rehabilitation. Unfortunately, existing systems come at a very high cost, certainly due to their high complexity. They are also generally quite cumbersome, also resulting in a limited usability, making it impossible to develop home rehabilitation. In the present project, we aim to make proofs of concept using disruptive solutions, based on emerging technologies. Our ambition is to pave the way to the development of a new generation of low-cost devices, inspired by a parsimonious approach for both the design and instrumentation, while in agreement with the medical needs.

• Co-PIs: B. Bayle (Unistra, ICube) & M. Gandolla (We-COBOT Polimi)

Development of multi-modality learning approaches for 3D medical imaging

Current AI-assisted diagnostic systems have shown success in triaging patients, generating reports, and improving clinical workflow. However, they require significant expert-based manual efforts for dataset generation to train the models and large-scale annotated datasets for successful generalization. This proposal aims to develop approaches for multi-modal self-supervised learning, working at the intersection of vision and language. Given radiological imaging and corresponding text reports, which are routinely collected in the clinical workflow, the goal is to learn a joint latent representation embodying rich semantic relationships between visual and textual features. The learned representations will be employed in various clinical applications such as automatic report generation, error detection, disease localization, and risk stratification. Our initial focus will be on medical imaging and text reports for laryngeal cancers, the second most prevalent form of head and neck cancer globally. We plan to expand to other types of cancers as we scale up. The proposed work packages include dataset generation, algorithm development, and performance evaluation.

We will collaborate with the All-India Institute of Medical Sciences (AIIMS), New Delhi, India, to provide clinical expertise to the project. Our clinical collaborators will help generate and curate the datasets and provide their expertise to evaluate the performance for downstream clinical applications.

• V. Srivastav, N. Padoy (IHU & ICube) and K. Rangarajan (All India Institute of Medical Sciences (AIIMS)

Curved trajectory planning solution for steerable needles with controlled stiffness 

The ARC team developed a needle with controlled stiffness/flexibility, enabling passive steering, relying on the locking, and unlocking of flexible joints at the needle’s distal end. Depending on the configuration, location, and number of joints defined, the needle’s flexibility can range from a rigid-like device to a very flexible needle. Thanks to its capabilities, three main use cases/indications are potentially enabled: non-linear trajectories, obstacle-avoidance, and real-time trajectory adjustments.

However, software-based solutions and user-friendly GUIs are needed to enable non-straight trajectory planning, and curvilinear needle insertions.

We aim to conceive, develop and validate a software-based solution to enable non-straight trajectory planning and insertion, using the ARC needle as an initial use case.

Co-PIs: C. Essert, L. Rubbert (ICube) and J. Verde (IHU)

A deep learning-based approach for real time acoustic simulations in therapeutic ultrasound

Recently, Focused Ultrasounds have demonstrated significant potential in brain therapies, such as neuromodulation and drug delivery. The correction of skull-induced aberrations, vital for precise targeting and optimal energy dosage, is currently a time-intensive process, impeding the integration
of these therapies into routine clinical practice. This project seeks to address this limitation by developing a novel tool capable of real-time computation of acoustic pressure maps within the brain so as to expedite and enhance the accuracy of treatment planning for physicians.

• co-PIs: P. Cabras, V. Srivastav, & N. Padoy, ICube & IHU CAMMA

Biomechanical modelling and numerical simulation of peripheral vertigo: a focus on various pathologies and patient specific surrogate

This project aims to consolidate diverse (2D-slice and 3D) existing anatomical finite element model of the vestibule of the human inner ear, and to use them to improve the knowledge of the physiology of common pathologies such as endolymphatic hydrops and benign paroxysmal positional vertigo. To
this end, various biomechanical models and numerical simulations are explored to lead to renewed diagnostic and therapeutic tools, this through a patient specific approach.

• co-PIs: D. Baumgartner & A. Charpoit, ICube & HUS

Biomechanical study of the distal radius resurfacing prosthesis: a comparative study of the Isoelastic and Cobra implants

When a patient experiences advanced wrist arthritis, surgery may be considered. In most cases, arthrodesis is performed, eliminating joint mobility. Another alternative is the wrist prosthesis; however it is often overlooked due to its low survival rate. The so called Isoelastic prothesis is known to provide better clinical outcomes. In this project, we propose a mechanical study of this prosthetic design based on its experimental comparison with the most used prosthesis model.

• co-PIs: P. Liverneaux & M. Bilasse, HUS & ICube
   

Characterization of polymer dental splints – Experimental focus on the risk of release microplastic

The potential ingestion of plastique particles in the case of polymer dental splints is a clinical and public health problem which is not enough analyzed. In this project, the potential release of particles will be quantified and analyzed on various types of splints using an innovative, patient-dependent,
experimental device to simulate chewing. This allows to build a database useful for a standard regarding particle release, damage and for better selection of splints.

• co-PIs: D. Wagner & N. Kharouf, ICube & INSERM 1121
   

Experimental measurement of the microstructure and mechanical properties of the Rectus Abdominis aponeurosis for the optimization of abdominal surgical stoma confection

The project focuses on abdominal parastomal hernia (PSH). To reduce the occurrence of PSH, better predictive numerical models are required. Unfortunately, there is a lack of experimental data. We aim to develop an adequate experimental protocol to provide both a precise knowledge of the
anisotropic structure and mechanical properties of the aponeurosis (or fascia) tissue which is the support of the prosthesis integrated by the surgeon to repair the PSH.

• co-PIs: D. George & B. Romain, ICube & INSERM U1113
   

Paradoxes in healthcare innovation: integrating grand challenges into innovation strategies

Grand challenges typically refer to significant and complex problems that require coordinated and innovative efforts to address. In the health industry, these grand challenges are even more important, as the integration of these challenges into organizational strategy has led organizations to evolve and
transform, incorporating new issues into their functioning, such as agility, collective intelligence, meaning at work, democratization, etc. (Gümüsay et al., 2022). Drawing on the dynamic capabilities theoretical framework, this research project aims to contribute insights into navigating paradoxes and
fostering responsible innovation in the healthcare industry, a major concern currently understudied.

• co-PIs: S. Bollinger, E. Ruiz, L. Leydiner, C. Dreyer, BETA & BioValley France
   

SILENCIO: Assessment of the impact of MRI acoustic noise on the time signature of brain networks

Over the past few decades, resting-state functional magnetic resonance imaging (fMRI) has attracted growing interest in monitoring brain network fluctuations in health and disease. Recent studies suggest that acoustic noise during fMRI acquisitions has an impact on the recording of neuronal
activity at rest. The aim of this project is to assess the impact of MRI acoustic noise on the temporal signature of networks using active noise reduction headsets.

• co-PIs: P. Louriero de Sousa & C. Meillier, ICube

Towards new tools to characterize complex post-traumatic stress disorder through an ecological, physiological and cognitive approach to emotional dysregulation

Complex posttraumatic stress disorder (cPTSD) is a new nosographic entity, internationally recognized since 2018. However, diagnosing and treating adults with cPTSD remain challenging since its symptoms lack specificity. This project aims to develop new tools and validate new methodologies
to investigate the clinical manifestations of cPTSD using a multi-level approach, combining ecological, physiological and cognitive measure in order to better characterize clinically cPTSD.

• co-PIs: A. Capobianco & L. Weiner, ICube & HUS
   

DAtabase for knowledge MOdelisation in RADiotherapy (DAMORAD)

Building structured radiotherapy databases of sufficient quality and quantity is particularly complex and time-consuming. This results in limited academic research, mostly based on a few dozen cases of learning data. This project aims to provide a model to facilitate the creation of massive radiotherapy databases for use in clinical and academic research projects.

• Co-PIs: P. Meyer & A. Lallement (ICANS & ICube)

 Mapping pH of tumors in vivo using CEST-MRI

The progressive reduction of extracellular pH within tumors, known as acidosis, contributes to multiple aspects of cancer progression. Therefore, quantification of tumor acidosis is essential for the determination of the therapeutic strategy, including radiotherapy planning. The main objective of this project is to evaluate the feasibility of MRI-CEST pH mapping in oncology.

• Co-PIs: P. Louriero de Sousa & C. Bund (ICube, ICANS)