Pakistan journal of medical sciences · 2025
Radiomics-Based Machine Learning in the Diagnosis of Type-B Aortic Dissection on Computed Tomography Images.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
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.
Shen Yifeng, Shi Xinyi, Ni Jianqi, Jin Qin, Wang Guoliang, Zou Jiajun
Аннотация
OBJECTIVE: To evaluate the value of a radiomics-based machine learning model in detecting Type-B aortic dissection (TBAD) on computed tomography (CT) images.
METHODOLOGY: This retrospective analysis included one hundred records of patients with clinically diagnosed TBAD and one hundred records of non-TBAD patients treated at the First Hospital of Jiaxing from January 2010 to January 2024. Radiomics features were extracted from CT non-contrast images, and the least absolute shrinkage and selection operator (LASSO) was used to construct dimensionality reduction and prediction models. The diagnostic performance of the model was evaluated through receiver operating characteristic (ROC) curves.
RESULTS: Fifteen radiomics features were extracted from the training cohort. All eight machine learning-established radiomics models in the training cohort demonstrated good prediction accuracy, with area under the ROC curve (AUC) values exceeding 0.9 in the validation set. Among the three models compared, the AUC values of the nomogram were the highest in both the training and validation cohorts (0.991 [95% confidence interval (CI): 0.982-1.000] and 0.998 [95% CI: 0.993-1.000], respectively). The calibration curves of the nomogram in both cohorts were more closely aligned with the dashed line. The nomogram showed the highest clinical benefits in both training and validation cohorts.
CONCLUSIONS: The predictive model established based on radiomics analysis of CT images demonstrates good predictive ability in recognizing TBAD.
METHODOLOGY: This retrospective analysis included one hundred records of patients with clinically diagnosed TBAD and one hundred records of non-TBAD patients treated at the First Hospital of Jiaxing from January 2010 to January 2024. Radiomics features were extracted from CT non-contrast images, and the least absolute shrinkage and selection operator (LASSO) was used to construct dimensionality reduction and prediction models. The diagnostic performance of the model was evaluated through receiver operating characteristic (ROC) curves.
RESULTS: Fifteen radiomics features were extracted from the training cohort. All eight machine learning-established radiomics models in the training cohort demonstrated good prediction accuracy, with area under the ROC curve (AUC) values exceeding 0.9 in the validation set. Among the three models compared, the AUC values of the nomogram were the highest in both the training and validation cohorts (0.991 [95% confidence interval (CI): 0.982-1.000] and 0.998 [95% CI: 0.993-1.000], respectively). The calibration curves of the nomogram in both cohorts were more closely aligned with the dashed line. The nomogram showed the highest clinical benefits in both training and validation cohorts.
CONCLUSIONS: The predictive model established based on radiomics analysis of CT images demonstrates good predictive ability in recognizing TBAD.