Prediction of Alzheimer's Disease Based on Bidirectional LSTM

Abstract

Alzheimer's disease (AD) is a common disease in the elderly. It affects human life seriously and is difficult to cure, so if you can predict the occurrence of the disease and the development trend in advance, you can prevent or treat Alzheimer's disease as soon as possible. Mild cognitive impairment (MCI) is a syndrome that occurs in the preclinical phase of Alzheimer's disease (AD). It is a transitional state between normal aging and early AD and may be an early sign of AD. This article uses the basic information of the patient's neuropsychological test scale data, genetic data and tomographic data in first, six and twelve months as input data and bidirectional LSTM plus Attention mechanism as a model to obtain a three-dimensional model. The output of the vector is divided into normal (NL), mild cognitive impairment (MCI) and Alzheimer's disease (AD). The experimental results can predict the development of Alzheimer's disease (AD), and determined the model has a good performance.