Psychiatry research · 2026
Differentiating unipolar and bipolar depression using multi-task eye-movement features.
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.
Zhong Rongrong, Zhang Yuezhou, Wu Xiaohui, Cui Lvchun, Lin Jiansen, Lv Xue, Chen Jiayue, Chen Shen Tse, Folarin Amos, Dobson Richard Jb, Chen Jun, Fang Yiru
Аннотация
BACKGROUND: Differentiating bipolar depression (BPD) from unipolar depression (UPD) is clinically challenging due to symptom overlap. This study explores eye-movement differences between UPD, BPD, and healthy controls (HCs) using a multi-task eye-tracking approach.
METHODS: Eye-movement data were collected from 228 participants (60 UPD, 56 BPD, 112 HCs) across four tasks: fixation stability, free-viewing, visual search, and smooth pursuit. A total of 155 eye-movement features were extracted and analyzed using robust analysis of covariance (ANCOVA) and machine learning for classification.
RESULTS: Significant differences were identified in 53 features. BPD was characterized by shorter total, average, and first fixation durations across all fixation stability conditions. During free-viewing task, attention to happy images showed a graded decline (HC > UPD > BPD), while BPD exhibited increased fixation allocation to threatening images and reduced saccade velocity and amplitude to negative stimuli. UPD demonstrated reduced efficiency in processing happy faces in the visual search task. Machine learning achieved good discrimination (mean area under the receiver operating characteristic curve [AUC]: 0.78 for HC vs. UPD; 0.88 for HC vs. BPD; 0.87 for UPD vs. BPD), with key contributing features overlapping with those identified as significant in statistical analyses.
CONCLUSIONS: Eye-movement patterns reveal both shared and disorder-specific features in UPD and BPD, supporting the potential of eye-tracking as an objective and scalable tool for differentiating mood disorders.
METHODS: Eye-movement data were collected from 228 participants (60 UPD, 56 BPD, 112 HCs) across four tasks: fixation stability, free-viewing, visual search, and smooth pursuit. A total of 155 eye-movement features were extracted and analyzed using robust analysis of covariance (ANCOVA) and machine learning for classification.
RESULTS: Significant differences were identified in 53 features. BPD was characterized by shorter total, average, and first fixation durations across all fixation stability conditions. During free-viewing task, attention to happy images showed a graded decline (HC > UPD > BPD), while BPD exhibited increased fixation allocation to threatening images and reduced saccade velocity and amplitude to negative stimuli. UPD demonstrated reduced efficiency in processing happy faces in the visual search task. Machine learning achieved good discrimination (mean area under the receiver operating characteristic curve [AUC]: 0.78 for HC vs. UPD; 0.88 for HC vs. BPD; 0.87 for UPD vs. BPD), with key contributing features overlapping with those identified as significant in statistical analyses.
CONCLUSIONS: Eye-movement patterns reveal both shared and disorder-specific features in UPD and BPD, supporting the potential of eye-tracking as an objective and scalable tool for differentiating mood disorders.