Frontiers in oncology · 2026
Intranodular and perinodular radiomics features based on non-contrast CT to distinguish pulmonary cryptococcosis from lung adenocarcinoma: a two-center study.
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.
Deng Yanfang, Lin Lin, Deng Kaiji, Xie Liqing, Lai Faming, Zhong Suhua, Jiang Yunshan, Xue Yunjing
Аннотация
BACKGROUND: Distinguishing pulmonary cryptococcosis (PC) from lung adenocarcinoma (LAC) remains clinically challenging in practice. The purpose of this study was to investigate the utility of intranodular and perinodular radiomics features derived from non-contrast CT in differentiating PC from LAC.
MATERIALS AND METHODS: A total of 244 patients with PC and LAC from two centers were randomly divided into a training set and a testing set at a ratio of 7:3. Logistic regression analysis was used to establish the clinical model. Radiomics features were extracted from the lesions and lesion margins of 10 mm. Support vector machine (SVM) was used to construct the intranodular, perinodular, and combined radiomics models. The areas under the receiver operating characteristic curve (AUCs) and decision curve analysis (DCA) were employed to assess the diagnostic performance, while the DeLong test was applied for model comparisons.
RESULTS: The three radiomics models exhibited excellent diagnostic performance for identifying PC and LAC, with the combined radiomics model achieving the highest AUC value in both the training (AUC = 0.936, sensitivity = 0.838, specificity = 0.898, accuracy = 0.859) and testing sets (AUC = 0.922, sensitivity = 0.854, specificity = 0.808, accuracy = 0.892). In the testing set, the AUC of the combined radiomics model was significantly higher than that of the clinical model (p = 0.005), while no statistically significant difference was found when compared with the intranodular or the perinodular model (p > 0.05). The combined model outperformed the other three models according to the DCA in terms of net benefit.
CONCLUSION: A combined radiomics model integrating intranodular and perinodular features can effectively improve diagnostic accuracy in differentiating PC from LAC.
MATERIALS AND METHODS: A total of 244 patients with PC and LAC from two centers were randomly divided into a training set and a testing set at a ratio of 7:3. Logistic regression analysis was used to establish the clinical model. Radiomics features were extracted from the lesions and lesion margins of 10 mm. Support vector machine (SVM) was used to construct the intranodular, perinodular, and combined radiomics models. The areas under the receiver operating characteristic curve (AUCs) and decision curve analysis (DCA) were employed to assess the diagnostic performance, while the DeLong test was applied for model comparisons.
RESULTS: The three radiomics models exhibited excellent diagnostic performance for identifying PC and LAC, with the combined radiomics model achieving the highest AUC value in both the training (AUC = 0.936, sensitivity = 0.838, specificity = 0.898, accuracy = 0.859) and testing sets (AUC = 0.922, sensitivity = 0.854, specificity = 0.808, accuracy = 0.892). In the testing set, the AUC of the combined radiomics model was significantly higher than that of the clinical model (p = 0.005), while no statistically significant difference was found when compared with the intranodular or the perinodular model (p > 0.05). The combined model outperformed the other three models according to the DCA in terms of net benefit.
CONCLUSION: A combined radiomics model integrating intranodular and perinodular features can effectively improve diagnostic accuracy in differentiating PC from LAC.