Automated peripheral neuropathy assessment of diabetic patients using optical imaging and binary processing techniques

Abstract

Plantar sensory neuropathy (PSN) affects a large proportion of individuals suffering from type 2 diabetes. In order to avoid ulceration or other damage to patients' feet regular testing and assessment of PSN needs to be undertaken. Currently, the standard test involves a trained podiatrist visiting the patient and testing their feet manually with a simple hand-held nylon monofilament probe. This process is time consuming, requires special training and is prone to errors. Moreover, the number of PSN sufferers is increasing and has already reached such numbers as to make manual testing unfeasible. Hence, our research team is currently developing an automated PSN testing device that will ultimately be capable of reliably testing a patient at home, providing direct feedback while registering and communicating this information to a local health care practice. An initial investigation is presented into a novel approach to automatically identify the areas of interest on a given patient's foot via optical image processing. That is to say, we present a method to reliably select suitable test points on the plantar surface that correspond to those chosen by a trained podiatrist. Once these points have been ascertained they will be sent to a microcontroller based robotic mechanism that subsequently tests the plantar surface with a monofilament probe. The robotic actuator will apply the probe to the pressure points of plantar surface a required number of times. On each application the patients' response, or lack of, will be recorded to identify the insensitive region of the plantar surface. The system will effectively automate the traditional Semmes-Weinstein monofilament examination (SWME). A GUI will be developed to show the statistics about patient sensory neuropathy.