Evaluation of blood perfusion of the trapezius muscle with wavelet analysis of photoplethysmogram signal using neural network

Measurement of tissue blood perfusion has many applications in the prevention of pressure sores, muscle activity assessment and care of tissue blood perfusion during surgery. Photoplethysmography as a continuous measure for evaluation of blood perfusion in tissue is accepted by researchers. In this study a new method for assessment of blood perfusion to the tissue based on photoplethysmograph signal (PPG) is presented. Wavelengths of the PPG were near infrared 950 nm with source-to-detector separation of 7 and 22 mm. The probe was placed over the trapezius muscle of 19 healthy subjects under the external pressures of 0 and 40 and 80 mmHg. PPG envelope detected and wavelet transform calculated in the five frequency intervals. These bands relate to metabolic, neurogenic, myogenic, respiratory and cardiac activities. The p-value of the t-test analysis for extracted features was less than 0.005. Results have shown that by applying external pressure, tissue deep layers most affected and the amount of their blood perfusion is reduced. Accuracy of separation at different pressures for back propagation neural network (BPNN) was 73.68% and for generalized regression neural network (GRNN) was 79.6%. Improvement of this method can be as a clinical assessment of tissue blood perfusion and can be as an effective method in prevention of pressure ulcers.