Oral surgery, oral medicine, oral pathology and oral radiology · 2026
Machine learning based on magnetic resonance imaging-derived texture features for differentiation of minor salivary gland tumors.
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.
Muraoka Hirotaka, Ito Kotaro, Otsuka Kohei, Komatsu Tomohiro
Аннотация
OBJECTIVE: We aimed to differentiate between benign and malignant minor salivary gland tumors using machine learning (ML) based on magnetic resonance imaging-derived texture features.
STUDY DESIGN: The study included 29 patients diagnosed with minor salivary gland tumors. To increase the effective dataset size for ML, data augmentation was applied, resulting in a total of 145 samples. The dataset included demographic variables (age, sex), MRI texture features, and lesion types. Age, sex, and the selected MRI texture features were used as predictor variables, while lesion type served as the outcome variable. The outcome variables were the types of minor salivary gland lesions. Multiple ML models-including Random Forest, logistic regression, Support Vector Machine, Extreme Gradient Boosting, and Linear Discriminant Analysis-were trained and evaluated using stratified 5-fold cross-validation. Classification performance was assessed using learning curves and confusion matrices.
RESULTS: ML analysis showed that the average precision, recall, and F1 scores were all between 0.80 and 0.97.
CONCLUSIONS: Our study demonstrated that ML using magnetic resonance imaging-derived texture features is a useful tool for the differentiation of benign and malignant minor salivary gland tumors.
STUDY DESIGN: The study included 29 patients diagnosed with minor salivary gland tumors. To increase the effective dataset size for ML, data augmentation was applied, resulting in a total of 145 samples. The dataset included demographic variables (age, sex), MRI texture features, and lesion types. Age, sex, and the selected MRI texture features were used as predictor variables, while lesion type served as the outcome variable. The outcome variables were the types of minor salivary gland lesions. Multiple ML models-including Random Forest, logistic regression, Support Vector Machine, Extreme Gradient Boosting, and Linear Discriminant Analysis-were trained and evaluated using stratified 5-fold cross-validation. Classification performance was assessed using learning curves and confusion matrices.
RESULTS: ML analysis showed that the average precision, recall, and F1 scores were all between 0.80 and 0.97.
CONCLUSIONS: Our study demonstrated that ML using magnetic resonance imaging-derived texture features is a useful tool for the differentiation of benign and malignant minor salivary gland tumors.