医学
案件调查
肌肉团
鉴定(生物学)
放射科
干预(咨询)
骨骼肌
体质指数
外科
计算机断层摄影术
作者
Ling Lee,Shu-Han Chuang,Yi‐Jie Kuo,Lien-Chen Wu,Yu‐Pin Chen
标识
DOI:10.1007/s41999-026-01465-3
摘要
Abstract Purpose Sarcopenia diagnosis requires identifying low muscle mass (LMM), typically via dual-energy X-ray absorptiometry (DXA). However, DXA's limited accessibility restricts large-scale screening. This retrospective study aimed to develop and validate a deep learning model to predict DXA-derived ASMI from routine hip radiographs for opportunistic sarcopenia screening. Methods We included 1,267 patients with both hip radiography and DXA scans, split into development (n = 1,140) and external validation (n = 127) sets. A multimodal model integrating radiographic images (ResNet-34 backbone) and clinical variables (age, gender, height, weight, BMI) was trained to predict continuous ASMI and classify LMM per Asian Working Group for Sarcopenia (AWGS) 2019 criteria. Results On external validation, the model achieved strong performance with Pearson r = 0.806, R²=0.650, MAE = 0.414 kg/m², and AUC = 0.874 for LMM classification. Applying AWGS diagnostic thresholds yielded sensitivity of 70.5% and specificity of 83.3%, with consistent performance across gender and age subgroups. Gradient-weighted Class Activation Mapping confirmed focus on clinically relevant gluteal and proximal thigh muscles. Conclusions This deep learning approach enables automated LMM identification from routine hip radiographs, offering a cost-effective, accessible tool for opportunistic sarcopenia screening and early intervention in at-risk populations.
科研通智能强力驱动
Strongly Powered by AbleSci AI