According to the statistic of World Health Organization, over 5% of the world’s population, i.e., 360 million people, has disabling hearing loss (328 million adults and 32 million children). Cochlear implant surgery can be used for profoundly deafened patient, for whom hearing aids are not satisfactory, and it is regarded as one of the best options for better hearing. During the implant surgery, the most difficult task is to insert the electrode array into the tympanic ramp of the patient's cochlea. The implant is normally made of silicone (thus very soft), the surgery is performed manually because the cochlear implant is totally passive and the surgeon has no perception on what happens in the cochlea while he/she is doing the insertion.

This thesis aims to significantly advance the automation of cochlear implant insertion, progressing from TRL 3 to TRL 6. It is partially funded by the ANR PRCE project ACCESS and seeks to address critical challenges in the modeling, simulation, and control of active Thin-Film Electroactive Actuators (TFEAs) for cochlear implantation. A primary focus of the research is to develop robust solutions for navigating the complex anatomy of the cochlea and its surrounding deformable structures, which present significant challenges for both the design of the implant and the precision of its insertion.

In our former ROBOCOP project, we developed modeling techniques for ECP and TFEA actuators, as well as control strategies for both passive and active cochlear implant insertions. However, to further enhance the level of robotization in implant insertion, it is crucial to consider not only the cochlea and the cochlear implant but also the surrounding anatomical structures, such as blood vessels, the facial nerve, and bones.

This requires:

• Modeling of soft tissues and deformable organs interacting with the active implant.
• Development of an innovative simulator that integrates the active implant, surrounding anatomical structures, and a robotic manipulator.

These new elements increase the complexity of trajectory planning and optimal control for insertions, necessitating a thorough reevaluation. The inclusion of anatomical structures imposes additional constraints, such as avoiding sensitive nerves.

This work will be conducted in close collaboration with IEMN (CNRS) for the modeling, the R&D department of Cochlea Company for the simulation, and the Institut de l’Audition (Institut Pasteur) for the validation. By combining modeling, simulation, real-time control strategies, and rigorous validation, it aims to contribute to the technological advancement of soft robotics in medical applications, particularly in the field of cochlear implantation.

Application Process: Please send your detailed CV, cover letter, and Bachelor’s and Master’s transcripts to: gang.zheng@inria.fr, christian.duriez@inria.fr, and yinoussa.adagolodjo@inria.fr

We aim to model the entire robotic system (manipulator and active implant), design an optimal control strategy, and collaborate to develop real-time simulations. This will enable the creation of an automated or semi-automated implant insertion process.

Main Tasks:

1. Modeling:

   • Develop multiphysics models integrating the dynamics of active TFEAs, the cochlea, and surrounding anatomical structures.
   • Employ advanced techniques such as Cosserat beam theory and finite element methods (FEM).
   • Generate patient-specific models using high-resolution CT scans.

2. Simulation:

   • Develop real-time numerical simulations with the SOFA framework to create digital twins of cochlear implantation.
   • Optimize implant geometry and plan surgical interventions.

3. Optimization and Control:

   • Design trajectory planning and closed-loop control strategies to avoid anatomical damage while ensuring precise implantation.

4. Validation:

   • Validate the models and simulations using clinical data and preclinical experiments.
   • Perform hardware tests and preclinical trials to evaluate the performance of the developed strategies.