Mild cognitive impairment disease classifying method based on brain network and brain structure information

Abstract

The invention discloses a mild cognitive impairment disease classifying method based on a brain network and brain structure information, and belongs to the field of medical image disease classification. The method includes the steps that individual cortex characteristics are calculated first, and difference between individual cortices is calculated with a specific kernel function to construct the structural brain network; any number of parameter types are selected from network characteristics and original MRI data to be combined, characteristics are selected from each combined datum through a stepwise judgment method, and the top ten characteristics appearing the most frequently are used as classification characteristics; a support vector machine based on a radial basis function is adopted for classification. The interference on characteristic selection by different combining modes of the cortex characteristics and network characteristics is fully considered, and robustness and stationarity of characteristic selection are ensured. The characteristics with high sensitivity can be selected fast and effectively, and classification precision is improved.

Description

technical field [0001] The invention belongs to the field of medical imaging disease classification, and in particular relates to a mild cognitive impairment disease classification method based on brain network and brain structure information. Background technique [0002] Alzheimer's disease (AD), manifested as severe memory loss and cognitive and behavioral impairment, is a genetically irreversible mental disorder, and it is the most common type of dementia. Mild Cognitive Impairment (MCI) is an intermediate stage between healthy aging and dementia. Risk of transferring to AD. As an intermediate stage of transition from normal aging to dementia, MCI has received extensive attention. According to the longitudinal diagnosis status of MCI patients, MCI can be divided into MCI patients who transform into AD patients within a certain period of time (MCIc) and patients with AD within a certain period of time. MCI patients (MCInc) who have not transformed into AD patients. Thr...

AI Technical Summary

Problems solved by technology

It can be seen that even if other types of data are added, the classifier does not have high resolution ability

At the same time, other modal data, such as the carrying number of the gene ApoEε4 related to the disease, the content and location of the specific protein in the cerebrospinal fluid related to the disease, and the related indicators of the patient's neuropsychology, etc., have a certain certainty when acquired. difficulty

Images