Brain function connection biomarker identification method based on neighborhood decision rough set

Abstract

The invention provides a brain function connection biomarker recognition method based on a neighborhood decision rough set, and relates to the two fields of brain science and the neighborhood decision rough set. the neighborhood decision rough set capable of effectively processing continuous and high-noise data is introduced for the first time aiming at the characteristics of continuity and high noise of the brain function connection data to improve the accuracy of brain function connection biomarker recognition; the efficiency of the neighborhood decision rough set for identifying the brain function connection biomarker is ensured by reducing a feature search space and quickly generating equivalence classes, namely, the coarse-grained feature search space of the neighborhood decision rough set is generated by grouping brain function connection features by using feature separability measurement; and the equivalence class of each sample is quickly obtained by using the symmetry of the Hash mapping and the neighborhood relationship. The method can accurately and rapidly obtain brain function connection discrimination characteristics with strong classification capability, and is expected to provide more accurate biomarkers for diagnosis of neurological and psychiatric diseases.

Description

technical field [0001] The present invention relates to the two fields of brain science and neighborhood decision-making rough sets. Aiming at the task of identifying the discriminative features of brain functional connectivity to obtain biomarkers of brain functional connectivity, a brain functional connectivity biomarker based on neighborhood decision-making rough sets is designed. object recognition method. Background technique [0002] Depression, Alzheimer's disease, autism and other neuropsychiatric diseases are a class of major diseases that endanger human health. At present, the pathological mechanism of most of these diseases is still unclear, and there is a lack of accurate and reliable biomarkers. As a result, the diagnosis mainly depends on the experience of doctors, which is full of subjectivity and cannot guarantee the objectivity and accuracy of diagnosis. Recent brain science research has found that many neuropsychiatric diseases are associated with abnormal...

AI Technical Summary

Problems solved by technology

But they only focus on the importance of a single feature, ignoring the correlation between features

In addition, the above two types of existing methods lack in-depth analysis of brain functional connectivity data, and it is difficult to obtain brain functional connectivity features with strong classification and discrimination capabilities.

Images