2024-2025

Drivers and Barriers to AI and Robotics Adoption in Medical Scienc

This project aims to study the barriers and drivers to the adoption of AI and robotics in the medical field. By focusing on the roles of local skills and collaboration, this project aims to map the ecosystem surrounding these technologies in the healthcare sector. It will be based on quantitative analyses of patent and scientific publication databases, as well as interviews with surgeons, computer scientists, and roboticists.

 

  • Supervisors: P. Pelletier (BETA), V. Schaeffer (BETA) & F. Nageotte (ICube)
  • HealthTech Master's student: Brian Aboujaoudé (PolyMtl)

Automatic detection and classification of carpal dislocations from frontal radiographs

 

Carpal dislocations are a rare but serious pathology, undetected about 25% of the time on initial exam. Delayed diagnosis gives rise to complications such as reduced function, pain, and degenerative arthritis. The goal of this project is to further develop a Deep Learning-based algorithm recently proposed by the supervisors for the automatic detection of perilunate dislocations in frontal radiographs, including robustness evaluation and improvement, as well as severity grading.

  • Supervisors: R. Allègre, P. Liverneaux, C. Wemmert
  • HealthTech Master's student: Mirna Alnakri

Enhanced Animal Wellbeing and Preclinical Research Quality with AI Monitoring (WELL-AIM)

This project aims to improve animal wellbeing in preclinical research by integrating advanced AI-driven monitoring systems. These systems provide comprehensive real-time tracking and analysis of animal health and behavior, aligning with the 3Rs principles of Replacement, Reduction, and Refinement. Continuous monitoring and early identification of adverse events ensure higher data quality and improve animal welfare, contributing to scientific rigor and humane practices, thereby advancing the principles of the 3Rs.

  • Supervisors: F. Wanert (IHU), N. Padoy (ICube), J. Verde (IHU)
  • HealthTech Master's student: Florent Dell'Aniello Picard (PolyMtl)

DENT-IA: Development of explainable artificial intelligence tools for the identification and classification of oro-dental anomalies based on multimodal data

DENT-IA project aims to use a neurosymbolic hybrid AI model, combining knowledge graphs and neural networks on multimodal data (clinical reports, imaging, genetics...) to improve the diagnosis and classification of rare dental diseases. This innovative approach merges expert knowledge with
data-driven learning to enhance diagnostic accuracy and development of personalized medicine for rare oro-dental diseases, hence promoting innovation in HealthTech through explainable AI.

  • Supervisors: A. Ayadi (ICube), C. Wemmert (ICube), O. Poch (ICube), A. Bloch-Zupan (IGMBC)
  • HealthTech Master's student: Kareem Elgohary

Advancing Multi-Modality Learning Methods for Automated Report Generation from 3D Head and Neck Cancer Medical Images

 

As per the latest GLOBOCAN estimates (2020), Head and Neck cancer is the seventh most common cancer globally, accounting for an estimated 890,000 new cases (roughly 4.5% of all cancer diagnoses around the world) and 450,000 deaths per year (roughly 4.6% of global cancer deaths. The growing utilization of medical imaging is increasing the burden on radiologists, as the time-consuming nature
of preparing radiology reports has become a hurdle in providing fast and efficient report generation from the 3D medical scan. This is especially relevant in the case of head and neck cancers given the growing number of cases worldwide. In this proposal, we aim to develop deep learning-based multi-modal solutions for automating report generation for head and neck cancers.

  • Supervisors: V. Srivastav & N. Padoy (IHU-ICube RDH CAMMA), K. Rangarajan (AIIMS, India)
  • HealthTech Master's student: Ayush Gupta

Biomechanical modeling and numerical simulation of peripheral vertigo: a focus on patient specific surrogate

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

  • Supervisors: D. Baumgartner & A. Charpiot - ICube
  • HealthTech Master's student: Louis Joly

Cognitive and emotional state detection in interactive time based on physiological sensors and behavioral data

 

Cognitive behavioral therapy helps people understand and regulate their own behavior, notably by being exposed to emotion-inducing situations through virtual reality. To ensure that the expected emotions are indeed conveyed, the main tool is self-assessment, unreliable and usable only after the situation. To propose virtual environment reacting to the user’s emotions, we wish to design a classifier allowing to detect emotional states in interactive time.

  • Supervisors: A. Capobianco, A. Ayadi - ICube
  • HealthTech Master's student: Moustafa Ahmed

Biomechanical study of connected prostheses

Myoelectric hand prostheses are designed to be ergonomic and lightweight using 3D printing. The current project aims to create an affordable prosthesis inspired by biomimicry, replicating the human hand's composition with composite skeleton structures and motorized cables for muscle simulation. The entire prosthesis is encased in silicone for accurate replication of the hand's shape. Inventive design methods and advanced composite materials will ensure the reliability of the prosthesis.

  • Supervisors: H. Chibane (INSA) & N. Bahlouli - ICube, P. Liverneaux - HUS
  • HealthTech Master's student: Margherita Nada

Developing physics-aware deep-learning approaches for real-time acoustic simulation for blood brain barrier opening planning

 

Focused ultrasound (FUS) can temporarily, safely and non-invasively open the blood-brain barrier (BBB) to deliver drugs that would not otherwise penetrate brain tissue. These procedures require planning stages based on lengthy multiphysics simulations. This project aims to explore the synergy between physical models and deep learning approaches to calculate acoustic pressure maps in the brain in real time, to help physicians plan the operation quickly, accurately and safely.

  • Supervisors: P. Cabras, V. Srivastav, N. Padoy - IHU-ICube RDH CAMMA
  • HealthTech Master's student: Juliette Puel

Multi-localization for the Digital Twin of a Hospital

Digital Twins (DTs) are promising tools to improve the management of Emergency Departments (EDs). Within a collaboration with the ED of Hautepierre (Strasbourg University Hospital), we need to provide a localization service to track in real-time the location of devices/patients/staff. The master student will develop a prototype to feed a DT of the Hautepierre ED, with a fusion of multiple different wireless technologies to improve indoor accuracy. Such an approach is the first step for a context-aware DT, benefiting from both models and measurements. 

  • Supervisors: F. Theoleyre & V. Goepp (INSA) - ICube
  • HealthTech Master's student: Luna Salameh

Design and Control of a Haptic Interface for Collaborative Needle Insertion Training

Interventional radiology is a practice that remains less developed than surgery, despite its potential to offer non-surgical alternatives for severe conditions. This project aims to provide a specific medical training device for percutaneous procedures. The system emulates a needle insertion procedure using a haptic interface and X-ray visual feedback. Finally, the haptic feedback provides assistance and guidance to the learner, either automatically or from a mentor through collaborative haptic training.

  • Supervisors: H. Omran, F. Nageotte & B. Bayle - ICube
  • HealthTech Master's student: Carlos Simoes

Focused Ultrasound-triggered Disruption of Bacterial Biofilms

Resistance of bacteria to antibiotics has been identified by the WHO as one of the top 10 health threats. This project aims to develop a new therapeutic approach using focused ultrasound. Indeed, recent studies suggest that the combination between ultrasound, microbubbles and antibiotics increases significantly the therapeutic antibacterial effect. The objective of this Master Research project is to develop a portable therapeutic ultrasound device and protocol and to evaluate it in vitro.

  • Supervisors: J. Vappou (ICube), G. Mislin (BSC)
  • HealthTech Master's student: Anna Soliani

Slice-to-volume stitching for free-hand oblique plane microscopy

Developing 3D tissue imaging techniques with fluorescence contrast would bring novel biomedical information, in particular for living tissues. Oblique Plane Microscopy (OPM) is a light-sheet technique that allows three-dimensional fluorescence imaging. It can be applied to tissue or cell imaging, issued from pre-clinical models or patients. The aim of this internship is to develop the volumetric reconstruction from 2D images series for free-hand or imperfect acquisition.

  • Supervisors: V. Maioli & D. Fortun - ICube
  • HealthTech Master's student: Anna Beatriz de Souza Perotto

Spectroscopic approach to identify tissue types

Raman spectroscopy has been applied to analyze a wide array of material across various fields. A Raman spectrum can provide the content of a material by relfecting a structural fingerprint. Compared to traditional histopathological analysis, tissue analysis using Rama spectroscopy is non-invasive and highly accurate, however could have low signal intensities due to its inherent low probability of occurence. This Master's thesis aims to use a Raman microscopy setup to identify tissue types by collecting a database from various tissue types.

  • Supervisor: F. Canbaz, University of Basel
  • HealthTech Master's student: Aurélien Guzzetti
    first international exchange for an internship in the framework of the SEMP mobility agreement signed in 2024 with the Biomedical Engineering Department at UniBas