Utilizing Previously Acquired BSBL Algorithm Parameters in the Compressed Sensing Framework for EEG Measurements

Compressed sensing (CS) has garnered significant attention for its potential to minimize power consumption in electroencephalogram (EEG) measurement devices. However, CS often requires substantial computational time for signal reconstruction. In this study, we introduce a novel approach aimed at reducing the reconstruction time in CS. We achieve this by reusing parameters obtained during the previous reconstruction as initial parameters for subsequent reconstruction processes. This method accelerates signal reconstruction without compromising accuracy. In Python3 simulations, our approach reduced computation time by a factor of 1.7. These findings provide valuable insights for designing low-power CS-based wireless EEG measurement systems.