BMC neurology · 2025
Characteristics of brain glucose metabolism in Parkinson's disease patients with freezing of gait: a study based on 18F-FDG PET imaging and deep learning.
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.
Zhu Zhuang, Geng Yao, Wang Xixi, Shi Jiaxin, Wang Hualin, Liu Linghui, Li Shengrong, Gan Caiting, Yuan Yongsheng, Zhu Qi, Zhang Kezhong
Аннотация
OBJECTIVE: Freezing of gait (FOG) is a common gait disorder in the advanced stages of Parkinson's disease (PD), closely associated with impaired balance and executive function. This study aimed to investigate specific changes in brain glucose metabolism in FOG patients using 18F-FDG PET. Deep learning methods were utilized to offer valuable perspectives for identifying FOG.
METHODS: Eighteen PD patients with FOG(PD-FOG), 11 patients without FOG (PD-NFOG) and 17 healthy controls (HC) were recruited. All participants underwent 18F-FDG PET imaging, and group comparisons were employed, to identify regions with significant differences in glucose metabolism. 3D convolutional neural network (3D CNN), as well as traditional machine learning models, were constructed for the automatic identification of the FOG type.
RESULTS: PET imaging analysis showed that the differences between the PD-FOG group and the PD-NFOG group were mainly located in the frontal lobe, parietal lobe and cingulate gyrus. The 3D CNN achieved diagnostic accuracies of 90.09% for distinguishing PD and 95.40% for FOG, surpassing other machine learning models. The 3D CNN achieved the smallest mean squared error (MSE), amounting to 48.01, in the prediction of Freezing of Gait Questionnaire (FOG-Q) scores.
CONCLUSION: Specific glucose metabolism patterns in PD-FOG mainly covered the frontoparietal network (FPN). The integration of 18F-FDG PET imaging with deep learning methods effectively differentiated patients with FOG. The 3D CNN exhibited a high diagnosis accuracy level, providing reliable imaging and artificial intelligence support for PD with FOG.
METHODS: Eighteen PD patients with FOG(PD-FOG), 11 patients without FOG (PD-NFOG) and 17 healthy controls (HC) were recruited. All participants underwent 18F-FDG PET imaging, and group comparisons were employed, to identify regions with significant differences in glucose metabolism. 3D convolutional neural network (3D CNN), as well as traditional machine learning models, were constructed for the automatic identification of the FOG type.
RESULTS: PET imaging analysis showed that the differences between the PD-FOG group and the PD-NFOG group were mainly located in the frontal lobe, parietal lobe and cingulate gyrus. The 3D CNN achieved diagnostic accuracies of 90.09% for distinguishing PD and 95.40% for FOG, surpassing other machine learning models. The 3D CNN achieved the smallest mean squared error (MSE), amounting to 48.01, in the prediction of Freezing of Gait Questionnaire (FOG-Q) scores.
CONCLUSION: Specific glucose metabolism patterns in PD-FOG mainly covered the frontoparietal network (FPN). The integration of 18F-FDG PET imaging with deep learning methods effectively differentiated patients with FOG. The 3D CNN exhibited a high diagnosis accuracy level, providing reliable imaging and artificial intelligence support for PD with FOG.