Annonce

Les commentaires sont clos.

Postdoctoral position at CEA Grenoble in ML for Brain Machine Interface

6 Mai 2022


Catégorie : Post-doctorant


Neural signal decoding for clinical Brain Spine neuroprosthesis

CEA/LETI/CLINATEC invite applications to work on the HORIZON-EIC project REVERSE PARALYSIS. This interdisciplinary project aims to develop and to test neuroprosthetics to restore limb movements in humans with chronic paralysis. Neuroprosthetics record, and decode brain neuronal signal to activate an effector directly without physiological neural control command pass way, which is interrupted. In the project, the neuronal activity is recorded at the level of cerebral cortex (ElectroCorticoGrams, ECoG) using chronic WIMAGINE implant. The ECoG data are decoded to send the command to an effector. A set of decoding algorithms were developed at CLINATEC, and were applied in clinical research protocols in Grenoble, and in Lausanne (STIMO-BSI clinical trial). The implantable spinal cord stimulator was used as an effector in STIMO-BSI to restore locomotion in paraplegic suffering from complete or partial spinal cord injury. Recent technological breakthroughs in cervical stimulation for restoring the upper limbs raises the new hope for rehabilitation of patients with chronic paralysis. Potential increase of the number of degrees of freedom raises new challenges for neuronal signal decoding. To support the development, CLINATEC is looking for a post-doctoral fellow / research engineer in machine learning and real time signal processing.

Context:

CEA/LETI/CLINATEC invite applications to work on the HORIZON-EIC project “Brain-Spine Interfaces (BSI) to Reverse Upper- and Lower-Limb Paralysis” (REVERSE PARALYSIS). REVERSE PARALYSIS is an interdisciplinary project, which aims to develop and to test neuroprosthetics to restore limb movements in humans with chronic paralysis. Neuroprosthetics record, and decode brain neuronal signal to activate an effector directly without physiological neural control command pass way interrupted by e.g. spinal cord injury. In the project, the neuronal activity is recorded at the level of cerebral cortex (ElectroCorticoGrams, ECoG) using chronic WIMAGINE implant. The signal is then decoded to send the command to the effector. A set of decoding algorithms were developed at CLINATEC [1], and were applied in clinical research protocols in tetraplegics in Grenoble (‘BCI&Tetraplegia’ clinical trial), and in paraplegic in Lausanne (STIMO-BSI clinical trial). The ‘BCI&Tetraplegia’ explores the solutions for tetraplegic patient functional compensation by controlling an exoskeleton. High-dimensional control of upper limbs is achieved [1], [2]. The implantable spinal cord stimulator was used as an effector in STIMO-BSI to restore locomotion in paraplegic suffering from complete or partial spinal cord injury [3]. Recent technological breakthroughs in cervical stimulation for restoring the upper limbs [4] raises the new hope for upper- and lower-limb rehabilitation of patients with chronic paralysis (REVERSE PARALYSIS project). Potential increase of the number of degrees of freedom raises the new challenges for neuronal signal decoding.

To support the development, CLINATEC is looking for a post-doctoral fellow in machine learning and neural signal processing. The postdoctoral fellowship will be carried out at the research technological center CEA (Commissariat à l’Energie Atomique et aux Energies Alternatives) at CEA/LETI/CLINATEC research institute (Grenoble, France), in collaboration with EPFL (Lausanne, Switzerland).

Mission:

Within a multidisciplinary team of clinicians and technologists, the postdoctoral fellow (M/W) will work in the field of machine learning / neural signal processing. She/he will contribute to the development of the algorithms for neuronal activity decoding and the upper limb(s) control performing real-life tasks. The prediction and control strategies will be developed / tested controlling different effectors and combinations: arm-hand exoskeleton, arm exoskeleton – hand FES (Functional Electrical Stimulation), and, finally, spinal cord stimulator. Cartesian control and direct joint control will be explored performing real-life tasks. Incremental / adaptive in real time machine learning algorithms will be applied for closed loop decoder training translating neuronal signal into effector control. Hidden semi-Markov Model will be employed to take into account the action temporal sequences performing real-life tasks. This will increase the decoding performance, responsiveness and robustness. Integration of developed algorithms in the real-time software platforms will allow testing of developed neuroprosthetics.

The postdoctoral fellow will be integrated into the signal processing team of CLINATEC. Participating in highly interdisciplinary project, he/she will collaborate with software and electrical engineers, biologists and medical doctors of both CLINATEC and EPFL teams.

[1] Moly, A., Costecalde, T., Martel, F., Martin, M., Larzabal, C., Karakas, S., ... & Aksenova, T. (2022). An adaptive closed-loop ECoG decoder for long-term and stable bimanual control of an exoskeleton by a tetraplegic. Journal of Neural Engineering, 19(2), 026021.

[2] Benabid, A. L., Costecalde, T., Eliseyev, A., Charvet, G., Verney, A., Karakas, S., ... & Chabardes, S. (2019). An exoskeleton controlled by an epidural wireless brain–machine interface in a tetraplegic patient: a proof-of-concept demonstration. The Lancet Neurology, 18(12), 1112-1122.

[3] Rowald, A., Komi, S., Demesmaeker, R., Baaklini, E., Hernandez-Charpak, S. D., Paoles, E., ... & Courtine, G. (2022). Activity-dependent spinal cord neuromodulation rapidly restores trunk and leg motor functions after complete paralysis. Nature Medicine, 1-12.

[4] Capogrosso, M., Barra, B., Conti, S., Perich, M., Zhuang, K., Schiavone, G., ... & Courtine, G. (2020). Cervical Epidural Electrical Stimulation Restores Voluntary Arm Control In Paralyzed Monkeys.

Applicant’s profile :

CLINATEC is looking for a post-doctoral fellow specializing in machine learning. The skills in adaptive / incremental learning, real-time processing of large data flows, as well as, the knowledge in deep learning will be appreciated. Experience in neural signal processing (EEG-ECoG-MEG) and strong interest in neuroscience and biomedical fields will be an advantage.

Date to start: preferable starting date is September 2022.

Application : CV, motivational letter and contact information for at least two referent scientists should be sent to Dr. Tetiana Aksenova, tetiana.aksenova@cea.fr.