Adaptive network-based fuzzy inference system for assessment of lower limb peripheral vascular occlusive disease

Detecting lower limb peripheral vascular occlusive disease (PVOD) early is important for patients to prevent disabling claudication, ischaemic rest pain and gangrene. According to previous research, the pulse timing and shape distortion characteristics of photoplethysmography (PPG) signals tend to increase with disease severity and calibrated amplitude decreases with vascular diseases. However, this is not a reliable method of evaluating the condition of PVOD because of noise effect. In this paper, an adaptive network-based fuzzy inference system (ANFIS) is proposed to assess lower limb PVOD based on PPG signals. PPG signals are non-invasively recorded from the right and left sides at the big toe sites from twenty subjects, including normal condition (Nor), lower-grade disease (LG), and higher-grade disease (HG) groups. The number of each group is 10, 8 and 2 respectively, and the ages ranged from 24 to 65 years. With the time-domain technique, the parameters for the absolute bilateral differences (right-to-left side of foot) in pulse delay and amplitude were extracted for analyzing ANFIS. The results indicated that ANFIS based on three timing parameters base bilateral differences, including ΔPTTf and ΔPTTp, and ΔRT has a high rate and noise tolerance of PVOD assessment.