Intermittent pneumatic compression (IPC) therapy has been adopted in prevention and treatment of ischemic-related peripheral vascular diseases. The aim of this study is to provide an approach to personalize the compression strategy of IPC therapy for maximizing foot skin blood flow. In this study, we presented a method to predict the optimized compression mode (OCM) for each subject based on biomechanical features extracted from experimental data tested with multiple IPC modes. First, to demonstrate the blood flow enhancing effect by applying the personalized OCM, four IPC modes of different frequency settings were tested on a total of 24 subjects. The frequency settings were adjusted by deflating-waiting time, which was defined as the total time length from the start of cuff deflation to the start of next compression. The foot skin blood perfusion and IPC air cuff pressure were monitored during the experiments. The personalized OCM was defined as the certain IPC mode that has the highest blood perfusion augmentation (BPA). Compared with the rest stage blood perfusion, the personalized OCM settings resulted in >50% of augmentation for 75% of healthy subjects (maximum augmentation at 244%) and >20% augmentation for 75% of patients with diabetes (maximum augmentation at 180%). Second, for predicting the OCM, we establish a random forest model based on the features extracted from the experimental data. The binary classification resulted in acceptable prediction performance (AUC > 0.7). This study might inspire new IPC strategies for improving foot microcirculation.
Keywords: Biomechanical modeling; IPC; Machine learning; Peripheral vascular disease; Personalized therapy.
Copyright © 2023 Elsevier Ltd. All rights reserved.