Development and validation of a clinical-radiomics nomogram for differentiating Mycoplasma pneumoniae pneumonia from bacterial pneumonia in children - Report - MDSpire
Advertisement
Development and validation of a clinical-radiomics nomogram for differentiating Mycoplasma pneumoniae pneumonia from bacterial pneumonia in children
Clinical Report: Creation and assessment of a clinical-radiomics nomogram
Overview
This study developed a nomogram that integrates clinical indicators and CT-based radiomics to differentiate Mycoplasma pneumoniae pneumonia from bacterial pneumonia in pediatric patients.
Background
Mycoplasma pneumoniae is a significant pathogen in pediatric community-acquired pneumonia (CAP), particularly among school-aged children. Differentiating Mycoplasma pneumoniae pneumonia (MPP) from bacterial pneumonia (BP) is clinically challenging due to overlapping symptoms and imaging features. Accurate differentiation is crucial for guiding appropriate antimicrobial therapy and improving patient outcomes.
Data Highlights
Model
Training AUC
Validation AUC
Clinical Model
0.913
0.909
Radiomics Model
0.918
0.895
Combined Nomogram Model
0.971
0.958
Key Findings
The study included 585 pediatric pneumonia patients, with 249 diagnosed with MPP and 336 with BP.
The clinical model achieved AUCs of 0.913 and 0.909 in training and validation sets, respectively.
The radiomics model reached AUCs of 0.918 and 0.895.
The combined nomogram model provided the highest accuracy with AUCs of 0.971 and 0.958.
Calibration curves confirmed the accuracy of the combined model.
Decision curve analysis indicated significant net clinical benefit from the combined model.
Clinical Implications
The combined nomogram model may assist clinicians in differentiating between MPP and BP in pediatric patients.
Conclusion
The development of a clinical-radiomics nomogram represents an advancement in the differential diagnosis of pneumonia in children.
