A Single-Chip Solution for Diagnosing Peripheral Arterial Disease

Peripheral arterial disease (PAD) is a common form of cardiovascular disease. The study proposes a noninvasive investigation of PAD based on the Doppler spectrogram of lower limb arteries. The proposed method consists of spectrogram image binarization using Otsu’s thresholding, feature extraction, and automated diagnosis using a support vector machine (SVM)-based classifier. The entire system is implemented in a field-programmable gate array with a target of wearable ultrasound (US) technology. The logarithmic domain-based approximate implementation reduces power consumption and design complexity without affecting performance significantly. Overall, the binary classification accuracy is found to be 90.40% in the study of 125 spectrograms. The back-end design can be useful to integrate with the US system for a cost-effective solution in the resource-constrained platform as well as point-of-care (POC) applications.