A vision transformer-convolutional neural network framework for decision-transparent dual-energy X-ray absorptiometry recommendations using chest low-dose CT

Duen-Pang Kuo, Yung-Chieh Chen, Sho-Jen Cheng, Kevin Li-Chun Hsieh, Yi-Tien Li, Po-Chih Kuo, Yung-Chun Chang, Cheng-Yu Chen

Abstract: Highlights•A ViT–CNN model outperforms individual models in bone classification using chest LDCT.•Thoracic FOV size affects model performance in bone classification.•A visualized DXA recommendation supports decision-making in clinical practice.•The framework improves accuracy and applicability in osteoporosis screening.