Deep-learning-based diagnosis framework for ankle-brachial index defined peripheral arterial disease of lower extremity wound: Comparison with physicians

• Published in: Computer methods and programs in biomedicine Vol. 263; p. 108654
• Main Authors: Tsai, Ming-Feng, Chu, Yu-Chang, Yao, Wen-Teng, Yu, Chia-Meng, Chen, Yu-Fan, Huang, Shu-Tien, Liu, Liong-Rung, Chiu, Lang-Hua, Lin, Yueh-Hung, Yang, Chin-Yi, Ho, Kung-Chen, Yu, Chieh-Ming, Huang, Wen-Chen, Ou, Sheng-Yun, Tung, Kwang-Yi, Hung, Fei-Hung, Chiu, Hung-Wen
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.05.2025
• Subjects: Algorithms; Ankle Brachial Index; Convolutional neural network; Deep Learning; Diagnosis, Computer-Assisted - methods; Female; Humans; Lower Extremity; Lower extremity wound; Male; Neural Networks, Computer; Peripheral Arterial Disease - diagnosis; Peripheral Arterial Disease - diagnostic imaging; Peripheral arterial occlusion disease; Peripheral artery disease; Physicians; Peripheral artery disease; Lower extremity wound; Ankle-brachial index; Convolutional neural network; Peripheral arterial occlusion disease
• ISSN: 0169-2607; 1872-7565; 1872-7565
• DOI: 10.1016/j.cmpb.2025.108654

•ResNet101V2-deep-learning is a great performance for PAD detection/classification.•Plastic and reconstructive surgeons and 3 models have a similar performance.•General practitioners show a lower performance in PAD detection and classification. Few studies have evaluated peripheral artery disease (PAD) in patients with lower extremity wounds by a convolutional neural network (CNN)-based deep learning algorithm. We aimed to establish a framework for PAD detection, peripheral arterial occlusive disease (PAOD) detection, and PAD classification in patients with lower extremity wounds by the AlexNet, GoogleNet, and ResNet101V2 algorithms. Our proposed framework was based on a CNN-based AlexNet, GoogleNet, or ResNet 101V2 model devoted to performing optimized detection and classification of PAD in patients with lower extremity wounds. We also evaluated the performance of the plastic and reconstructive surgeons (PRS) and general practitioner (GP). Compared to the performance of AlexNet or GoogleNet, a slight increase in ResNet101V2-based performance of PAD detection, PAOD detection, and PAD classification with original images was observed. A similar observation was found for PAD detection, PAOD detection, and PAD classification with background-removal or cropped images. GP group had a lower performance for PAD and PAOD detection than did the three models with original images, while a similar performance for PAD detection was observed in PRS group and the 3 models. We proposed a promising framework using CNN-based deep learning based on objective ankle-brachial index (ABI) values and image preprocessing to characterize PAD detection, PAOD detection, and PAD classification for lower extremity wounds, which provides an easily implemented and objective and reliable computational tool for physicians to automatically identify and classify PAD.