Neural network application to the development of a novel diabetic neuropathy diagnosis tool using the valsalva index and the SCR

Diabetes mellitus is a chronic metabolic disease that displays hyperglycaemia and that is strongly linked to micro and macro-vascular complications and neuropathic ones. The World Health Organization (WHO) states that there are around 171 million diabetic patients in the world, it´s also estimated that this amount will double by 2030. We have performed a preliminary study on 35 volunteers, including diabetic patients and controls, whose results suggest that the skin conductance response (SCR) measured during the Valsalva maneuver could be used in the development of a diagnosis index of diabetic neuropathy. The phasic component and the Valsalva index of each patient were used as input for a neural network to identify patients with diabetic neuropathy. Preliminary results indicate that the index obtained through the neural network is indeed a viable approach to a fast subject classification.