Adaptive multi-band electroencephalogram signal mild cognitive impairment assessment method

• Main Authors: NI TONGGUANG, QIAN PENGJIANG, GU XIAOQING, JIANG YIZHANG, XUE JING
• Format: Patent
• Language: Chinese; English
• Published: 28.12.2021
• Subjects: DIAGNOSIS; HUMAN NECESSITIES; HYGIENE; IDENTIFICATION; MEDICAL OR VETERINARY SCIENCE; SURGERY

The invention relates to the technical field of cognitive impairment, in particular to an adaptive multi-band electroencephalogram signal mild cognitive impairment assessment method, which comprises the following steps of: 1, presenting standardized sensory stimulation to a subject; 2, acquiring electroencephalogram signals of the subject; 3, performing feature extraction on the electroencephalogram signals according to frequency bands; 4, calculating a set of nearest neighbor similar samples and nearest neighbor heterogeneous samples of each sample; 5, constructing a similar association graph matrix; 6, constructing a heterogeneous association graph matrix; 7, performing modeling based on a relative entropy and a subspace adaptive learning model; and 8, classifying test samples. According to the invention, on the basis of the concept of the relative entropy, it can be guaranteed that space structure information of a sample is kept in the dimensionality reduction process; the electroencephalogram signals are