Frontiers in pediatrics · 2026
Development and validation of a clinical-radiomics nomogram for differentiating Mycoplasma pneumoniae pneumonia from bacterial pneumonia in children.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Guan Yan, Wang Xueqin, Song Chen, Bi Lulin, Yang Guang, Quan Shuai, Xu Shuming
Аннотация
OBJECTIVE: In this case-control study, we developed a nomogram merging computed tomography (CT)-based radiomics, clinical indicators, and CT imaging findings for differentiating Mycoplasma pneumoniae Pneumonia (MPP) from bacterial pneumonia (BP) in children.
METHODS: We retrospectively analyzed clinical and CT imaging data from 585 pediatric pneumonia patients, including 249 with MPP and 336 with BP. Patients were randomly allocated to training (70%) and validation (30%) groups. CT images were segmented using PHIgo-LK segmentation software (GE Healthcare), and radiomics features were extracted. The minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) were used to screen the key features in the training group and the corresponding radiomics score were obtained. We developed three models: clinical, radiomics, and a combined nomogram model. Model performance was evaluated using the receiver operating characteristic (ROC) curve, calibration curves, and decision curve analysis.
RESULTS: The clinical model (variables: white blood cell count, C-reactive protein, lactate dehydrogenase, tree-fog sign, and bilateral lesions) achieved areas under the receiver operating characteristic curve (AUC) of 0.913 and 0.909 in the training and validation sets, respectively. The radiomics model built from five selected features reached AUCs of 0.918 and 0.895. Integration of clinical variables, CT morphology, and radiomics score into a nomogram delivered the highest accuracy, with AUCs of 0.971 and 0.958. Calibration curves confirmed the model's accuracy, and decision curve analysis highlighted significant net clinical benefit.
CONCLUSION: The combined nomogram model could provide a decision-making basis for early clinical differentiation of MPP from BP in children.
METHODS: We retrospectively analyzed clinical and CT imaging data from 585 pediatric pneumonia patients, including 249 with MPP and 336 with BP. Patients were randomly allocated to training (70%) and validation (30%) groups. CT images were segmented using PHIgo-LK segmentation software (GE Healthcare), and radiomics features were extracted. The minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) were used to screen the key features in the training group and the corresponding radiomics score were obtained. We developed three models: clinical, radiomics, and a combined nomogram model. Model performance was evaluated using the receiver operating characteristic (ROC) curve, calibration curves, and decision curve analysis.
RESULTS: The clinical model (variables: white blood cell count, C-reactive protein, lactate dehydrogenase, tree-fog sign, and bilateral lesions) achieved areas under the receiver operating characteristic curve (AUC) of 0.913 and 0.909 in the training and validation sets, respectively. The radiomics model built from five selected features reached AUCs of 0.918 and 0.895. Integration of clinical variables, CT morphology, and radiomics score into a nomogram delivered the highest accuracy, with AUCs of 0.971 and 0.958. Calibration curves confirmed the model's accuracy, and decision curve analysis highlighted significant net clinical benefit.
CONCLUSION: The combined nomogram model could provide a decision-making basis for early clinical differentiation of MPP from BP in children.