Cortex; a journal devoted to the study of the nervous system and behavior · 2024
Data-driven prediction of regional brain metabolism using neuropsychological assessment in Alzheimer's disease and behavioral variant Frontotemporal dementia.
СтатьяПоддержка исследования: вне правительства СШАНазвание 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.
Díaz-Álvarez Josefa, García-Gutiérrez Fernando, Bueso-Inchausti Pedro, Cabrera-Martín María Nieves, Delgado-Alonso Cristina, Delgado-Alvarez Alfonso, Diez-Cirarda Maria, Valls-Carbo Adrian, Fernández-Romero Lucia, Valles-Salgado Maria, Dauden-Oñate Paloma, Matías-Guiu Jorge, Peña-Casanova Jordi, Ayala José L, Matias-Guiu Jordi A
Аннотация
BACKGROUND: This study aimed to evaluate the capacity of neuropsychological assessment to predict the regional brain metabolism in a cohort of patients with amnestic Alzheimer's disease (AD) and behavioral variant frontotemporal dementia (bvFTD) using Machine Learning algorithms.
METHODS: We included 360 subjects, consisting of 186 patients with AD, 87 with bvFTD, and 87 cognitively healthy controls. All participants underwent a neuropsychological assessment using the Addenbrooke's Cognitive Examination and the Neuronorma battery, in addition to [18F]-fluorodeoxyglucose positron emission tomography (FDG-PET) imaging. We trained Machine Learning algorithms, including artificial neural networks (ANN) and models that incorporate genetic algorithms (GAs), to predict the presence of regional hypometabolism in FDG-PET imaging based on cognitive testing results.
RESULTS: The proposed models demonstrated the ability to predict hypometabolism trends with approximately 70% accuracy in key regions associated with AD and bvFTD. In addition, we showed that incorporating neuropsychological tests provided relevant information for predicting brain hypometabolism. The temporal lobe was the best-predicted region, followed by the parietal, frontal, and some areas in the occipital lobe. Diagnosis played a significant role in the estimation of hypometabolism, and several neuropsychological tests were identified as the most important predictors for different brain regions. In our experiments, classical Machine Learning models, such as support vector machines enhanced by a preliminary feature selection step using GAs outperformed ANNs.
CONCLUSIONS: A successful prediction of regional brain metabolism of patients with AD and bvFTD was achieved based on the results of neuropsychological examination and Machine Learning algorithms. These findings support the neurobiological validity of neuropsychological examination and the feasibility of a topographical diagnosis in patients with neurodegenerative disorders.
METHODS: We included 360 subjects, consisting of 186 patients with AD, 87 with bvFTD, and 87 cognitively healthy controls. All participants underwent a neuropsychological assessment using the Addenbrooke's Cognitive Examination and the Neuronorma battery, in addition to [18F]-fluorodeoxyglucose positron emission tomography (FDG-PET) imaging. We trained Machine Learning algorithms, including artificial neural networks (ANN) and models that incorporate genetic algorithms (GAs), to predict the presence of regional hypometabolism in FDG-PET imaging based on cognitive testing results.
RESULTS: The proposed models demonstrated the ability to predict hypometabolism trends with approximately 70% accuracy in key regions associated with AD and bvFTD. In addition, we showed that incorporating neuropsychological tests provided relevant information for predicting brain hypometabolism. The temporal lobe was the best-predicted region, followed by the parietal, frontal, and some areas in the occipital lobe. Diagnosis played a significant role in the estimation of hypometabolism, and several neuropsychological tests were identified as the most important predictors for different brain regions. In our experiments, classical Machine Learning models, such as support vector machines enhanced by a preliminary feature selection step using GAs outperformed ANNs.
CONCLUSIONS: A successful prediction of regional brain metabolism of patients with AD and bvFTD was achieved based on the results of neuropsychological examination and Machine Learning algorithms. These findings support the neurobiological validity of neuropsychological examination and the feasibility of a topographical diagnosis in patients with neurodegenerative disorders.