Abstract:It has always been a big challenge to identify subtle changes in Electroencephalogram (EEG) signals. Minor differences often lead to vital decisions, for example, which grade a certain tumour belong to or whether a haemorrhage can result in benign blood clots or cancerous ones. In recent studies on brain computer interfaces (BCIs), one of the biggest challenges is recovering maximum information for realistic predictions. In order to choose EEG channels with highest accuracy, a novel notion of including sparsity in a modified common spatial pattern (CSP) algorithm is introduced here. Being influenced by the existing concept of compressed sensing, an optimization model is also developed alongside to recover the cosparse signal and retain maximum information. The state-of-the-art Joint Sparsity Induced Modified Common Spatial Pattern Algorithm and Low Rank Optimization Model (SCSP-LROM) developed here is capable of identifying and describing tumours and lesions in great detail at an overall accuracy of 96.3%.

Joint SCSP-LROM: A novel approach to detect Cerebrovascular Anomalies from EEG signals