A CT-based deep learning model for the automated risk stratification of refractory Mycoplasma pneumoniae pneumonia in children.
Authors
Affiliations (15)
Affiliations (15)
- Department of Radiology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China.
- 4+4 Medical Doctor Program, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China.
- Theranostics and Translational Research Center, National Infrastructures for Translational Medicine, Institute of Clinical Medicine, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China.
- National Clinical Research Center for Child Health, Children's Hospital, National Children's Regional Medical Center, Zhejiang University School of Medicine, Hangzhou, 310052, China.
- Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China.
- Department of Data and Information, Children's Hospital, Zhejiang University School of Medicine, Hangzhou, China.
- Department of Evidence-Based Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China.
- Department of CT, Rizhao Hospital of Traditional Chinese Medicine, Rizhao, China.
- Department of Pediatrics, Rizhao Hospital of Traditional Chinese Medicine, Rizhao, China.
- Deepwise AI Lab, Beijing Deepwise & League of PhD Technology Co.Ltd, Beijing, China.
- Department of Radiology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China. [email protected].
- Department of Radiology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, No. 1 Shuaifuyuan, Dongcheng District, Beijing, 100730, People's Republic of China. [email protected].
- National Clinical Research Center for Child Health, Children's Hospital, National Children's Regional Medical Center, Zhejiang University School of Medicine, Hangzhou, 310052, China. [email protected].
- Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, Hangzhou, China. [email protected].
- National Clinical Research Center for Child Health, Children's Hospital, Sino-Finland Joint AI Laboratory for Child Health of Zhejiang Province, National Children's Regional Medical Center, Zhejiang University School of Medicine, No. 3333 Binsheng Road, Binjiang District, Hangzhou, 310000, People's Republic Of China. [email protected].
Abstract
The accurate identification of children with refractory Mycoplasma pneumoniae pneumonia (RMPP) remains challenging. This study aimed to develop a transformer-based model utilizing clinically indicated chest computed tomography (CT) to stratify pediatric RMPP risk at a critical decision point. Non-contrast chest CT data from a multicenter retrospective cohort of 1224 pediatric patients with Mycoplasma pneumoniae pneumonia who underwent clinically indicated CT were used to develop a transformer-based deep learning framework (trans-DLF). The primary cohort comprised training (n = 506), validation (n = 140), and internal testing (n = 139) cohorts, with two independent external cohorts (n = 331 and n = 108) used to evaluate generalizability. Model performance was assessed by the area under the receiver operating characteristic curve (AUC) and compared against a three-dimensional convolutional neural network (3D-CNN), a clinical model, and a multimodal nomogram. Interpretability was examined using gradient-weighted class activation mapping (Grad-CAM). The median age was 6.83 years (interquartile range, 5.0-8.6 years), and 609 (49.8%) were male. The trans-DLF demonstrated strong performance across all cohorts: training (AUC 0.97; 95% confidence interval [CI], 0.96-0.98), validation (0.91; 0.86-0.96), internal testing (0.90; 0.85-0.95), and external testing (0.89; 0.84-0.94 and 0.89; 0.82-0.95). It significantly outperformed the clinical model (p < 0.001), while its AUCs were not significantly different from those of the multimodal nomogram. The model maintained good performance in outpatient settings (AUC 0.87) with good calibration and net clinical benefit. Grad-CAM suggested that predictions were influenced by clinically meaningful features, particularly consolidations. The trans-DLF provides a streamlined and efficient approach to RMPP risk assessment in children who have already undergone clinically indicated chest CT and may support timely, evidence-based decision-making without additional tests.