Studies On The Mental States Classification Based On The Resting-State Functional Connectivity

In recent years,with the development of medical science and other sciences,the recognition and treatment of mental illness has made great progress.Specially,during the development of MRI techniques,using fMRI to study mental illness has become a hot spot.However,many problems remain unsolved.For example,the criteria for determining mental illness are not accurate enough.So,we want to quantify mental state using the fMRI data of healthy people.On the one hand,it can be used with the criteria to help doctor give accurate judgement.On the other hand,it's easier to judge the extent to which individuals deviate from a good mental state,and then give appropriate advice and treatment to different levels of individuals to help them develop into a good mental state.To verify the feasibility of quantification,we need to discuss the differences between healthy people in different mental states and the differences between healthy people and psychopath,and to judge whether the two differences is consistency in brain activity.Since the differences between healthy people and psychopath have been studied by many researchers,the main task of us is to find out the differences between healthy people in different mental states.The object is to be identified as the undiseased group by the DSM-IV,collectively known as the healthy person.Since the research object has no label,we use statistics,machine learning,brain science and other related knowledge to design algorithms to solve the problem of unsupervised machine learning.Through the algorithm,several groups of depression related and different mental states were isolated.We found that there were some common differences between these groups in brain activity.By comparing with the differences between healthy people and depression patient,both have a high degree of overlap.The results showed that the consistency is founded.This provides a theoretical basis for the quantification of mental state based on the fMRI data of healthy people.Characteristics:1.Due to the strong subjectivity of the existing judgment standard,we use restingstate fMRI to find the differences in brain activity of healthy people with different mental states,which lay the foundation for quantifying mental state.2.Since most studies focus only on the sick and undiseased groups,we broaden people's understanding of mental illness from the perspective of healthy people.3.This paper provides a new way to solve the problem of unsupervised machine learning.