EEG Artifact Removal

The observations of EEG signals are typically corrupted by many physiological effects and electronic/environmental perturbations. The removal of theses artifacts facilitates the physician in proper interpretation of the brain signals. The objective of the proposed research is to explore and identify the best among the existing techniques for artifact removal in EEG. One of the objectives of the project is to develop a real-time artifact removal system that will be employed in our indigenous EEG machines.


The competitive and evolving market trends imply that more and more technological input goes into analyzing how the consumers perceive different products and their promotions. This analysis is useful for the companies to adapt their product promotions in line with the user preferences. Neuromarketing is a recent concept that combines signals from the brain along with the some other markers to analyze response patterns of the users subjected to a visual stimulus. In this regard, EEG sensors allow acquisition of brain signals which are then correlated to other markers for analysis.

P300 and Odd-Ball Paradigm

P300 is an event related potential that is generated by nervous system in response to stimulus.P300 is generated during process of decision making. P300 has its applications in Brain Computer interfacing. P300 speller is type of BCI which uses EEG signals and P300 response evoked by visual stimuli in order to select an item on the computer screen. The goal of this research is to determine whether correct item is selected or not and send feedback for correction if wrong selection has been made.

Classification of Multi-Class Motor Imagery EEG

Brain Computer Interface (BCI) systems enable users to direct commands to an electronic device by using only their brain signals. Motor Imagery (MI) is a technique used to operate BCIs, in which communication with an external device is performed by composing a sequence of mental tasks such as imagination of movement of left and right hand, tongue, or feet. The objective of the project was to find an optimized algorithm for efficient and accurate classification of multi-class MI based EEG signals using limited number of channels.

NeuroInformatics Portal

Data acquisition and sharing is one of the prime hurdles in the progress of neuroinformatics research. Another problem is the access to high computational power that is required to process the huge amount of data. This project endeavors to create a cluster of computational resources available at NUST-SEECS, housing the open-source neural analysis software tools and publically available datasets, providing access to other labs and hospitals who lack this resource. This will allow other researchers to submit their code online and see the results without downloading any data or without requiring any computational strength.

Wireless EEG

Ambulatory EEG is beginning to find uses in many clinical and healthcare domains, entertainment, gaming, and broader brain computer interfaces. It allows for recording of neural signals outside the controlled environment of a lab in freely moving patients and subjects. With a renewed focus on health and emergence of fitness gadgets, wireless recording of EEG has found a niche in the non-medical marketplace as well. We endeavor to make a decently reliable wireless EEG acquisition system.

Automatic Seizure Detection

Epileptic patients suffer from recurrent and unpredictable seizures due to neural disorders in brain. This study focuses on automatic detection of epileptic seizure in Pediatric patients, a great challenge due to high variability in EEG data. The aim of this study is real time classification of electrical activity of brain.

Neuro-Imaging based Diagnostics

Brain-imaging provides a lot of information about functional and structural links between different parts of the brain, and has been used extensively for detecting various diseases at different stages of progression. Functional imaging has been extensively used to detect and monitor cognitive diseases. However, early detection of diseases, and related therapeutics, are still areas needing more research and improvement, and form the nucleus of our study.

Multimodal Neuro-Imaging for Cognition

Neuroimaging techniques (EEG, MEG and FMRI) measure distinct and complementary information related to neuronal activity in brain. The main application areas for neuroimaging are in structural and functional mapping of brain, brain-computer interfacing, and understanding neuronal activity of brain diseases. In this research we will devise an algorithm which improves the detection accuracy of perceptual tasks across multiple subjects using multimodal functional neuroimaging dataset.

Neural Spike Detection for Implantable Brain Circuits

With the advancement of technology, present-day multimodal intracranial recording systems offer high temporal and spatial resolution needed for brain machine interface systems. Real time on-chip spike detection is the first step in decoding neural spike trains in implantable brain machine interface systems.