RETRACTED: Optimizing the impact of time domain segmentation techniques on upper limb EMG decoding using multimodal...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Neurological disorders cause widespread motor impairments.
• Existing electromyography (EMG) decoding research focuses on classifiers and features, neglecting preprocessing techniques.
• Time-domain windowing in EMG signal preprocessing is an understudied area.

Purpose of the Study:

• To investigate the impact of different time-domain windowing techniques on electromyography (EMG) decoding accuracy for upper limb movements.
• To identify the most effective windowing technique for classifying finger movements.
• To bridge the knowledge gap regarding EMG preprocessing in motor function studies.

Main Methods:

• Recorded surface EMG data from volunteers performing fifteen distinct finger movements.
• Compared the performance of eight different time-domain windowing techniques.
• Utilized 40 time-domain features and a Linear Support Vector Machine (L-SVM) classifier for movement classification.

Main Results:

• The rectangular window technique demonstrated superior performance among the tested methods.
• Achieved a classification accuracy of 99.98% using the rectangular window, 40 time-domain features, and an L-SVM classifier.
• High accuracy confirms the efficacy of surface EMG for precise upper limb movement classification.

Conclusions:

• The rectangular windowing technique is highly effective for EMG-based upper limb movement classification.
• Optimized EMG preprocessing can significantly enhance the accuracy and reliability of myoelectric control systems.
• Findings support advancements in prosthetic limbs, wearable sensors, human-computer interaction, and brain-computer interfaces.