Multi Signal Pulse Wave Analysis for the Identification of Vascular Diseases Leading to Diabetic Foot

Abstract

The peripheral arterial diseases are usually diagnosed by non-invasive investigations, such as hemodynamic assessment of lower level arterial circulation, tissue oxygen perfusion measurements etc. Diabetic foot ulceration is mostly related with peripheral arterial diseases. This work aims to identify the possibilities of diabetic foot by analysing the differential pulse waves of arm and leg. Pulse volume waveforms of ankle and brachium are used for the analysis. Multi signal packet wavelet feature extraction technique is used for identifying the differential features of the right/left limbs. Machine learning classification algorithms are employed for the evaluation purpose. Previous studies have proved the direct relationship between pulse wave velocity and blood pressure. Pulse wave velocity is very much linked with vascular diseases. The Moens Korteweg and Hughes models provides the background for this study, which relates the blood pressure and Pulse Wave Velocity. Samples are collected form the normal diabetic patients and from those who have considerable symptoms of arterial diseases. Multi signal packet wavelet feature extraction techniques are used for identifying the differential features of the right/left limbs. Machine learning classification algorithms are used to identify the accuracy of the method.