Physical activity classification in A-ECG signals using neuro-fuzzy classifiers

Wearable ambulatory ECG (A-ECG) signals obtained using wearable ECG recorders inherently contain the motion artifacts due to various physicals activities of the subject. Classification of four such physical activities (PAs) - left arm up-down, right arm up-down, waist twisting and walking- of five healthy subjects has been performed using neuro-fuzzy classifier (NFC). The Gabor energy feature vectors have been used to train the NFC. The overall PA classification accuracy achieved by the NFC classifier is almost 95% for single-fold as well as ten-fold experiments.