BMC cancer · 2025
Multi-modal radiomics features to predict overall survival of locally advanced esophageal cancer after definitive chemoradiotherapy.
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.
Yu Nuo, Wan Yidong, Zuo Lijing, Cao Ying, Qu Dong, Liu Wenyang, Deng Lei, Zhang Tao, Wang Wenqing, Wang Jianyang, Lv Jima, Xiao Zefen, Feng Qinfu, Zhou Zongmei, Bi Nan, Niu Tianye, Wang Xin
Аннотация
PURPOSE: To establish prediction models to predict 2-year overall survival (OS) and stratify patients with different risks based on radiomics features extracted from magnetic resonance imaging (MRI) and computed tomography (CT) before definite chemoradiotherapy (dCRT) in locally advanced esophageal squamous cell carcinoma (ESCC).
METHODS: Patients with locally advanced ESCC were recruited. We extracted 547 radiomics features from MRI and CT images. The least absolute shrinkage and selection operator (LASSO) for COX algorithm was used to obtain features highly correlated with survival outcomes in the training cohort. Based on MRI, CT, and the hybrid image data, three prediction models were built. The predictive performance of the radiomics models was evaluated in the training cohort and verified in the validation cohort using AUC values.
RESULTS: A total of 192 patients were included and randomized into the training and validation cohorts. In predicting 2-year OS, the AUCs of the CT-based model were 0.733 and 0.654 for the training and validation sets. The MRI radiomics-based model was observed with similar AUCs of 0.750 and 0.686 in the training and validation sets. The AUC values of hybrid model combining MRI and CT radiomics features in predicting 2-year OS were 0.792 and 0.715 in the training and validation cohorts. It showed significant differences in 2-year OS in the high-risk and low-risk groups divided by the best cutoff value in the hybrid radiomics-based model.
CONCLUSIONS: The hybrid radiomics-based model demontrated the best performance of predicting 2-year OS and can differentiate the high-risk and low-risk patients.
METHODS: Patients with locally advanced ESCC were recruited. We extracted 547 radiomics features from MRI and CT images. The least absolute shrinkage and selection operator (LASSO) for COX algorithm was used to obtain features highly correlated with survival outcomes in the training cohort. Based on MRI, CT, and the hybrid image data, three prediction models were built. The predictive performance of the radiomics models was evaluated in the training cohort and verified in the validation cohort using AUC values.
RESULTS: A total of 192 patients were included and randomized into the training and validation cohorts. In predicting 2-year OS, the AUCs of the CT-based model were 0.733 and 0.654 for the training and validation sets. The MRI radiomics-based model was observed with similar AUCs of 0.750 and 0.686 in the training and validation sets. The AUC values of hybrid model combining MRI and CT radiomics features in predicting 2-year OS were 0.792 and 0.715 in the training and validation cohorts. It showed significant differences in 2-year OS in the high-risk and low-risk groups divided by the best cutoff value in the hybrid radiomics-based model.
CONCLUSIONS: The hybrid radiomics-based model demontrated the best performance of predicting 2-year OS and can differentiate the high-risk and low-risk patients.