PhD (M/FF) in robotics and neurosciences

Job Description

Thesis Subject

**Title:** Biofeedback-Based Training Approaches for Learning Selective Activation of Arm Muscle Heads for Prosthesis Control

**Methodological bottleneck of the thesis:** Despite recent scientific advances, people with arm amputations fitted with prostheses still lack sufficient voluntarily controllable muscle signals to control all the joints of current prosthetic devices. In practice, only the biceps and triceps are usable, which strongly limits the number of controllable degrees of freedom. The most widely used solution today is a major and complex surgical procedure, Targeted Muscle Reinnervation (TMR), which aims to reroute amputated nerves in order to artificially generate new muscular control sites.

However, our observations in some individuals with agenesis show that it is possible to independently contract different heads of the biceps and triceps, thereby increasing the number of available EMG signals without surgery. The scientific bottleneck is therefore to understand how this ability developed and to test whether it can be reproduced through voluntary learning. The methodological bottleneck is therefore to design and evaluate specific training protocols — targeted exercises, visual and somatosensory biofeedback — that could make this skill accessible to any amputee.

**Positioning in relation to the state of the art:** Myoelectric control of prostheses still relies mainly on EMG detection from two antagonist muscles, resulting in slow sequential control that is often abandoned by users (Biddiss & Chau, 2007). Studies have shown that the heads of the biceps and triceps display partially independent recruitment patterns depending on position and task, suggesting a potential for voluntary differentiation (Murray et al., 2000; Staudenmann et al., 2009). However, no study has yet aimed at systematic training for the selective activation of arm muscle heads.

Encouraging results exist with EMG biofeedback on other muscles, enabling more localized control after only a few learning sessions (Holtermann et al., 2010; Lendaro et al., 2018). Finally, our clinical observations of a case of transhumeral agenesis confirm that such differentiated control can be exploited to manage several prosthetic degrees of freedom. This project is therefore positioned at the interface between classical myoelectric control approaches and emerging strategies based on differentiated recruitment, with the aim of experimentally demonstrating the feasibility of such learning in non-amputated subjects.

**References:**

• Biddiss, E., & Chau, T. (2007). Upper limb prosthesis use and abandonment: A survey of the last 25 years. *Prosthetics and Orthotics International.*

• Murray, W.M., et al. (2000). Variation in muscle activation patterns at the elbow with changes in shoulder and forearm position. *Journal of Biomechanics.*

• Staudenmann, D., et al. (2009). Reproducible EMG patterns for muscle function estimation in human movement. *Journal of Electromyography and Kinesiology.*

• Holtermann, A., et al. (2010). Selective activation of muscle subregions with EMG biofeedback training. *Journal of Applied Physiology.*

• Lendaro, E., et al. (2018). Myoelectric training with real-time feedback improves muscle activation control. *Frontiers in Human Neuroscience.*

**Interdisciplinary nature of the thesis:** The scientific bottleneck of the project lies in the ability to transform a voluntary muscle signal, potentially differentiated at the level of muscle heads, into functional and intuitive prosthesis control. This challenge cannot be addressed without a strong connection between neuroscience and robotics.

Neuroscience is essential to understand neuromuscular plasticity, characterize the possibilities of differentiated recruitment, and design appropriate learning and biofeedback protocols. Robotics, in turn, is indispensable for developing EMG acquisition devices, signal-processing algorithms, and their integration into multi-degree-of-freedom prosthetic control architectures.

The validity of the project therefore relies on the complementarity between the detailed analysis of motor-control mechanisms and their translation into usable robotic solutions. Only a joint approach can make it possible to move from a physiological proof of concept to a concrete application for prosthesis users.

Your Work Environment

Team IRIS of ISIR (UMR7222) -Sorbonne Université
Institut des sciences du mouvement - Etienne-Jules Marey (ISM) (UMR7287) - Aix Marseille Uni.
The supervisors have been collaborating for many years on the topic of the phantom limb phenomenon in amputees through the ANR PhantomovControl project, and they are currently collaborating on these same questions within the framework of the PEPR O2R REINVENT project.

Constraints and risks

none

Compensation and benefits

Compensation

2300 € gross monthly

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions