Postdoctoral position: Adaptive Brain-Motor Mapping
Neuroprosthetics is an interdisciplinary field related to
neuroscience, bioelectronics and biomedical engineering, which aims to
substitute a motor, sensory or cognitive function that might have been
damaged as a result of an injury or a disease. One of the challenging
issues in motor prosthesis is the large variety of patient situations
depending on the type of neurological disorder. To overcome the current
limited performance of such systems, a robust bio-signal processing and
a model-based control taking about the actual sensory motor state with
biosignal feedback would bring a break-through and allow to progress
toward adaptive neuroprosthesis.
Recent advances of Brain-Computer-Interfaces (BCI) have opened a new
communication channel for patients, who can transmit their movement
intention via brain signals.
The functionality and controllability of motor prosthesis can be further
improved by taking advantage of computational mapping between EMG
(Electromyography) of muscle, EEG (Electroencephalography) of brain, and
other modalities of biofeedback information.
The first objective is to enhance the classification algorithm to
extract the subject's motor intention from EEG signal in motor-imagery
based BCI. The computational modeling between multichannel EMG and EEG
will involve advanced feature extraction, dimension reduction and
classification algorithms. Moreover EMG signals of multiple muscles and
muscle modeling including skeletal dynamics models will help in
obtaining the detailed motion intention of the subject.
The second objective is to develop a bilateral learning architecture. In
BCI, adaptive decoding of EEG signals is desirable because brain signals
change over time during the learning of the task. In motor control, it
is also known that we change how we use our joints and a way to deal
with redundancy problems in articulation. In EMG analysis, the change of
motor usage can be captured. Adaptive modeling of EMG allows the
evaluation of skill acquisition.
By jointly analyzing both the EEG and EMG modifications, we investigate
how EEG signal may change along with actual motor coordination changes.
By modeling both of these adaptive features, this framework will try to
capture the bilateral learning architecture of both the brain and the
This project is associated with INRIA OpenVibe project (http_nospam_openvibe.inria.fr/)
Required skills: A strong background in signal processing, control, and
machine learning is required. Fluency in English, and excellent
programming skills (C++ and Matlab) are necessary.
To start from Sep/Oct 2014, for 1-2 years.
To apply for this position, please contact
Maureen Clerc (http_nospam_www-sop.inria.fr/members/Maureen.Clerc/)
Mitsuhiro Hayashibe (http_nospam_www.lirmm.fr/~hayashibe/)
Forum for discussion on different signal processing algorithms
1 post • Page 1 of 1
Who is online
Users browsing this forum: No registered users and 1 guest