Design and Enactment Evaluation of Adaptive Artifacts Removal from EEG Signal Records

Various factors, such as electrical power lines, EOG or ECG interference, contribute to artefacts in Electroencephalogram (EEG) data, complicating EEG analysis and clinical interpretation. Developing specialized filters to mitigate these artefacts is crucial. Artefacts from eye movements and blinks have been extensively studied, prompting the development of an FLM optimization-based learning technique for a Neural Network (NN)-enhanced adaptive filtering model to address them. Initially, Firefly (FF) and LM adaptive filter algorithms analyze EEG data to determine optimal weights. These weights are then incorporated into the NN for adaptive filtering. The resulting technique effectively eliminates artefacts. Performance evaluation, based on Signal to Noise Ratio (SNR), Root Mean Square Error (RMSE), Mean Square Error (MSE), and computing time, compares the proposed method with conventional approaches. Results demonstrate a significant 92% improvement in SNR, indicating the efficiency of the proposed technique. This advancement holds promise for enhancing EEG data quality and facilitating more accurate clinical assessments.