Pediatric blood & cancer · 2026
Predicting Chronicity in Children and Adolescents With Newly Diagnosed Immune Thrombocytopenia at the Timepoint of Diagnosis Using Machine Learning-Based Approaches.
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.
Kasser Severin, Bizeul Alice, Chitlur Meera, Donato Hugo, Roganovic Jelena, E Vogt Julia, Kühne Thomas
Аннотация
OBJECTIVES: To identify predictors of chronic ITP (cITP) and to develop a model based on several machine learning (ML) methods to estimate the individual risk of chronicity at the timepoint of diagnosis.
METHODS: We analyzed a longitudinal cohort of 944 children enrolled in the Intercontinental Cooperative immune thrombocytopenia (ITP) Study Group (ICIS) Children's Initiative. cITP was defined as a platelet count <100 × 109/L at 12 months post diagnosis. Thirty-six clinical and laboratory variables collected at diagnosis were evaluated, and key predictors were selected using ML approaches. Model performance was assessed by the area under the receiver operating characteristic (ROC) curve.
RESULTS: cITP developed in 28.7% of patients. Six variables were identified as the most informative predictors of chronicity. The ML model achieved an area under the ROC curve of 73.7% for predicting cITP at diagnosis.
CONCLUSIONS: ML-based prediction models can identify children at increased risk for cITP at the time of diagnosis. Integration of such tools into prospective registries may enhance prognostic accuracy and support individualized clinical decision-making.
METHODS: We analyzed a longitudinal cohort of 944 children enrolled in the Intercontinental Cooperative immune thrombocytopenia (ITP) Study Group (ICIS) Children's Initiative. cITP was defined as a platelet count <100 × 109/L at 12 months post diagnosis. Thirty-six clinical and laboratory variables collected at diagnosis were evaluated, and key predictors were selected using ML approaches. Model performance was assessed by the area under the receiver operating characteristic (ROC) curve.
RESULTS: cITP developed in 28.7% of patients. Six variables were identified as the most informative predictors of chronicity. The ML model achieved an area under the ROC curve of 73.7% for predicting cITP at diagnosis.
CONCLUSIONS: ML-based prediction models can identify children at increased risk for cITP at the time of diagnosis. Integration of such tools into prospective registries may enhance prognostic accuracy and support individualized clinical decision-making.