Search

Biomedical signal processing is essential for diagnosing and monitoring various health disorders. Conventional signal processing methods frequently struggle to effectively deal with biomedical data's intricate, nonlinear, and non-stationary characteristics, such as electroencephalograms (EEG). Time-frequency analysis (TFA) approaches provide valuable insights into the temporal variations of signal components, offering a possible way to address these difficulties. This study aimed to present time-frequency approaches, their applications across various biomedical signals, challenges encountered, and potential future directions. To highlight the proposed concept, Epilepsy (mental disturbance) was selected. The epileptic datasets, which included 3570 pairs of electrooculography (EEG) signals, were utilized. These signals were categorized into focal and non-focal signals. By applying the proposed approach that involves a scalogram-short time Fourier transform as a feature enhancement then fed to the convolution neural network CNN, the model achieved accuracy, precision, recall, and f1-score for the proposed model 76%,78%,75%, and 76% respectively. The results depict that integrating multiple time-frequency analysis techniques enhances biomedical signal processing capabilities, offering significant improvements in diagnosing and monitoring health conditions. [ABSTRACT FROM AUTHOR]