Multi-task-based feature selection method for the functional brain network under multiple thresholds

Abstract

According to the multi-task-based feature selection method for the functional brain network under multiple thresholds, multi-level network features are extracted in a multi-threshold mode, and multi-level features are extracted from the thresholded network in a multi-core multi-task learning mode for further classification processing. The defects of an existing method are overcome, and then the characteristics with discrimination and interpretation are learned. According to the gk-MTFS method, feature learning under each threshold value is taken as a task, structured information of a network is reserved for each task by adopting a graph core (a core constructed on a graph), and internal correlation among tasks is explored by adopting multi-task learning, so that features with higher discrimination and interpretability are learned. Finally, a real brain disease data set is verified, and experimental results show that compared with a method at the present stage, the method provided by the invention has better classification characteristics on the brain diseases.

Description

technical field [0001] The invention belongs to the fields of machine learning and medical image analysis, and in particular relates to a multi-task-based feature selection method under multi-threshold facing functional brain networks. Background technique [0002] With the rapid development of biotechnology, brain imaging technology, such as modern magnetic resonance imaging (MRI) technology, including functional MRI (functional MRI, fMRI), provides a non-invasive way to explore the human body The brain, revealing previously unrecognized mechanisms of brain structure and function. Brain network analysis can describe the interaction between brain regions at the connection level, and has become a new research hotspot in medical image analysis and neuroimaging. [0003] Recently, machine learning methods have been used in the analysis and classification of brain networks. For example, researchers have used brain networks for the diagnosis and classification of early brain di...

AI Technical Summary

Problems solved by technology

However, on the one hand, there is no good standard to choose a specific threshold, on the other hand, different thresholds generally result in different network structures, which may contain complementary information, which may further improve network analysis performance

Images