Addictive behaviors · 2026
Inter-subject variability in brain connectivity predicts nicotine dependence severity and differentiates smokers via machine learning.
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.
Cui Shaoyu, Xu Xuefeng, Yang Bo, Zheng Yanbin, Li Xuzhou, Dong Guang-Heng
Аннотация
BACKGROUND: Tobacco use disorder (TUD) is a major public health issue with significant individual differences. A deeper understanding of its neurobiology and reliable biomarkers is needed. This study investigated whether inter-subject variability in resting-state brain functional connectivity (IVFC) could serve as such a marker for TUD.
METHODS: Resting-state fMRI data from 123 male TUD patients and 123 healthy controls (HCs) were collected and analyzed. IVFC was computed within seven major brain lobes. Five machine learning models (random forest, gradient boosting, extra trees, multi-layer perceptron, and support vector machine) were trained to classify the groups based on IVFC features. Univariate regression was conducted with each lobe's IVFC predicting clinical scores. Multivariate stepwise regression was then performed to identify the best combination of IVFC predictors for dependence severity.
RESULTS: TUD showed significantly altered IVFC in six brain lobes compared to controls, with the largest difference in the insular lobe. The machine learning models, particularly the extra trees classifier, achieved high accuracy (up to 88 %) in classifying TUD, primarily utilizing features from the temporal and limbic lobes. Univariate regression showed that higher IVFC in frontal, insular, limbic, and temporal lobes was associated with lower nicotine dependence severity. Multivariate stepwise regression identified insular lobe IVFC as the sole independent predictor.
CONCLUSIONS: These findings suggest that widespread alterations in IVFC may characterize TUD and relate to its clinical severity, indicating the potential value of this measure as a neurobiological marker. The combination of IVFC analysis with machine learning appears to offer a promising approach for distinguishing TUD, which could contribute to advancing our understanding of its underlying brain mechanisms.
METHODS: Resting-state fMRI data from 123 male TUD patients and 123 healthy controls (HCs) were collected and analyzed. IVFC was computed within seven major brain lobes. Five machine learning models (random forest, gradient boosting, extra trees, multi-layer perceptron, and support vector machine) were trained to classify the groups based on IVFC features. Univariate regression was conducted with each lobe's IVFC predicting clinical scores. Multivariate stepwise regression was then performed to identify the best combination of IVFC predictors for dependence severity.
RESULTS: TUD showed significantly altered IVFC in six brain lobes compared to controls, with the largest difference in the insular lobe. The machine learning models, particularly the extra trees classifier, achieved high accuracy (up to 88 %) in classifying TUD, primarily utilizing features from the temporal and limbic lobes. Univariate regression showed that higher IVFC in frontal, insular, limbic, and temporal lobes was associated with lower nicotine dependence severity. Multivariate stepwise regression identified insular lobe IVFC as the sole independent predictor.
CONCLUSIONS: These findings suggest that widespread alterations in IVFC may characterize TUD and relate to its clinical severity, indicating the potential value of this measure as a neurobiological marker. The combination of IVFC analysis with machine learning appears to offer a promising approach for distinguishing TUD, which could contribute to advancing our understanding of its underlying brain mechanisms.