Application of DCT-derived parameters for early detection of polyneuropathy in diabetic patients

Abstract

AbstractDiabetic foot is one of the most severe complications of diabetes. Early diagnosis of this syndrome can ensure proper medical care and adequate treatment. Various image analysis methods can be used to speed up the diagnosis process, and automated diagnosis can be applied as a screening technique to reduce its cost. Introducing auxiliary diagnostic parameters may help to detect polyneuropathy or neuropathy, both of which often precede the appearance of diabetic foot syndrome. The present paper describes a study performed on a group of diabetic patients by analyzing plantar pressure distribution images. As part of this study, 2D discrete cosine transform (DCT) is computed for the forefoot and rearfoot regions of each diabetic subject in a group of 37 patients. Three new DCT-based parameters are introduced to help to detect polyneuropathy or at least indicate that the patient may have polyneuropathy without a time-consuming examination. The results indicate a certain relationship between these parameters and the presence of polyneuropathy. This information could be used in further diagnosis to prevent foot ulcers from developing in patients with diabetes.