Digestive and liver disease : official journal of the Italian Society of Gastroenterology and the Italian Association for the Study of the Liver · 2025
Machine learning models for predicting chemotherapy-induced adverse drug reactions in colorectal cancer patients.
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.
Wu Yingfei, Zhao Wei, Zhang Linxiao, Wang Ying, Wen Ying, Liu Lan
Аннотация
BACKGROUND: Chemotherapy-induced adverse drug reactions (ADRs) are common in patients with colorectal cancer. We developed four machine learning models to predict chemotherapy-induced ADRs and assessed the performance. These models leverage high-dimensional data and non-linear relationships, offering better predictive accuracy than traditional methods.
METHODS AND MATERIALS: We used 11 variables (age, sex, number of chemotherapy cycles, hypertension, diabetes, chemotherapy regimen, liver and kidney function, history of adverse reactions, concomitant radiotherapy, and bevacizumab use) to predict six chemotherapy-induced ADRs: nausea and vomiting, neutropenia, thrombocytopenia, anemia, peripheral neuropathy, bleeding, and overall ADRs. Four models were assessed (Random Forest, Support Vector Machine, XGBoost and LASSO Regression).
RESULTS: A total of 4072 chemotherapy cycles were analyzed, including2,147 XELOX, 768 FOLFOX, 1005 FOLFIRI, and 152 other cycles. The accuracy of the models in predicting overall ADRs was 0.8779(Random Forest), 0.8377(Support Vector Machine), 0.8758(XGBoost) and 0.7443(LASSO Regression) respectively. The models predicted individual ADRs better than overall ADRs of which Random Forest performed the best.
CONCLUSION: These models effectively predict chemotherapy side effects, offering improved accuracy and generalization over traditional methods. They aid clinicians in making informed decisions and contribute to personalized medicine for patients.
METHODS AND MATERIALS: We used 11 variables (age, sex, number of chemotherapy cycles, hypertension, diabetes, chemotherapy regimen, liver and kidney function, history of adverse reactions, concomitant radiotherapy, and bevacizumab use) to predict six chemotherapy-induced ADRs: nausea and vomiting, neutropenia, thrombocytopenia, anemia, peripheral neuropathy, bleeding, and overall ADRs. Four models were assessed (Random Forest, Support Vector Machine, XGBoost and LASSO Regression).
RESULTS: A total of 4072 chemotherapy cycles were analyzed, including2,147 XELOX, 768 FOLFOX, 1005 FOLFIRI, and 152 other cycles. The accuracy of the models in predicting overall ADRs was 0.8779(Random Forest), 0.8377(Support Vector Machine), 0.8758(XGBoost) and 0.7443(LASSO Regression) respectively. The models predicted individual ADRs better than overall ADRs of which Random Forest performed the best.
CONCLUSION: These models effectively predict chemotherapy side effects, offering improved accuracy and generalization over traditional methods. They aid clinicians in making informed decisions and contribute to personalized medicine for patients.