By Noman Naseer
The Neurorobotics Research Group, headed by Dr. Noman Naseer, has been working on developing fNIRS-based brain-computer interfaces since 2015. The group is part of the Mechatronics Engineering Department at Air University, Islamabad, Pakistan.
Here we describe three ongoing fNIRS projects:
Design and Fabrication of CW-fNIRS system
This project intends to develop fNIRS sensory array for human brain signal acquisition. Software for signal filtering, feature extraction and classification for BCI purpose is being developed as well.
fNIRS-based control of upper prosthetic limb
Every year, people lose their ability to perform useful work due to loss of limbs. The goal of this project is to enable disabled individuals to control their upper limbs through brain signals acquired using fNIRS. We are currently developing a prosthetic arm capable of two directional motion and limited wrist movement. The acquired brain signals are first preprocessed to remove noise and these noise free signals are then classified to generate control commands for different movements of the prosthetic arm.
fNIRS-based Neurorobotic Interface for gait rehabilitation
In this project, fNIRS signals are used to initiate and stop the gait cycle, while a nonlinear proportional derivative computed torque controller (PD-CTC) with gravity compensation is used to control the torque of hip and knee joints for minimization of position error. The brain signals generated during walking intention and rest tasks are acquired from the left hemisphere’s primary motor cortex. Thereafter, for removal of motion artifacts and physiological noises, the performances of six different filters (i.e. Kalman, Wiener, Gaussian, hemodynamic response filter (hrf), Band-pass, finite impulse response) were evaluated. Then, six different features are extracted from oxygenated hemoglobin signals, and their different combinations were used for classification. Also, classification performance on five different classifiers (i.e. k-Nearest Neighbour, quadratic discriminant analysis, linear discriminant analysis (LDA), Naïve Bayes, support vector machine (SVM)) were tested. The control commands that generated using the classifiers initiate and stop the gait cycle of the prosthetic leg, the knee and hip torques of which are controlled using the PD-CTC to minimize the position error. The proposed scheme can be effectively used for neurofeedback training and rehabilitation of lower-limb amputees and paralyzed patients.