A collaboration between Associate Professor Teppei Matsui of the Faculty of Natural Science and Technology (Science and Biology), Okayama University, and Specially Appointed Assistant Professor Trung Quang Pham of the Institute of Physiology, Associate Professor Koji Jimura of Keio University (Science and Technology), and Senior Researcher Junichi Chikazoe of ArayaInc. The group used statistical modelling of brain activity to make a discovery that challenges widely held beliefs about the brain's default mode.

　The results of these studies will be published as a Research Article in the US neuroscience journal Neuroimage on 1 April (already available online).

　The human brain is active even when it is not doing anything, which is known as the brain's default mode. In this study, we used statistical modelling to test a method widely used to analyse the brain's default mode, and found problems with it. The results show that the widely held belief that the brain's default state consists of multiple stable states may be wrong.

　The results of this research are expected to be an important starting point for the application of the brain's default mode to the diagnosis of dementia and neuropsychiatric disorders.

＜Current status＞

　The human brain is active even when it is not doing anything, known as the brain's default mode. Previous research has shown that examining brain activity in the default mode can reveal information about intellectual ability, gender, mental illness and much more. However, the fundamental nature of the brain's default mode is still poorly understood. The widely accepted theory is that the brain's default mode has a number of stable states, which it switches between from time to time, just as our thoughts come and go when we are not doing anything. However, this theory has not been fully tested to see if it is actually true.

＜The results of the research＞

　A joint research group consisting of Associate Professor Matsui, Specially Appointed Assistant Professor Pham at the Institute of Physiology, National Institutes of Natural Sciences, Associate Professor Jimura at Keio University (Science and Technology), and Researcher Chikazoe at ArayaInc. has tested the theorem that there are multiple stable states in the default mode of the brain. For this reason, we decided to focus on co-activity pattern analysis (1), a typical method to visualize the stable state of the default mode, and to test its validity.
We first performed statistical modelling using brain activity data in the default mode published by the Human Connectome Project (2) in the USA. From this statistical modelling, we obtained pseudo data that had only one stable state, while maintaining the characteristics of the real brain activity data. We then applied co-activity pattern analysis to both the real and the pseudo data and compared them. We found that the spatial and temporal patterns obtained from the co-activity pattern analysis were almost identical between the real and the pseudo data. This indicates that the conventional interpretation that there are multiple stable states in the default mode of the brain is wrong.
　Rather, the results suggest that the brain's default mode may have only one stable state, contrary to the conventional view.

THE DEFAULT MODE OF THE BRAIN. THE CALLOUT IS A SNAPSHOT OF BRAIN ACTIVITY MEASURED BY FUNCTIONAL MRI. IN DEFAULT MODE, THE BRAIN TRANSITIONS BETWEEN SEVERAL STABLE STATES. IN THE OPPOSING HYPOTHESIS, THE APPARENTLY DIFFERENT PATTERNS OF BRAIN ACTIVITY ACTUALLY ARISE FROM A SINGLE STATE. THE PRESENT STUDY SUGGESTS THAT THE OPPOSING HYPOTHESIS MAY BE MORE LIKELY TO BE CORRECT.

＜Social significance＞

　The brain's default mode is expected to have applications in the diagnosis of dementia and mental disorders. We will continue to research the nature of the brain's default mode, as revealed in this study, for the development of diagnostic techniques.

Article Information
　Title of paper: On co-activation pattern analysis and non-stationarity of resting brain activity
　Paper：Neuroimage
　Author： Teppei Matsui, Trung Quang Pham, Koji Jimura, Junichi Chikazoe
　D O I: https://doi.org/10.1016/j.neuroimage.2022.118904
　U R L: https://www.sciencedirect.com/science/article/pii/S1053811922000349

Research Funding
This research was supported by the following research funds; Grant-in-Aid for Scientific Research on Innovative Areas（Brain information dynamics underlying multi-area interconnectivity and parallel processing, Project/Area Number: 20H05052）、日本医療研究開発機構AMED（「国際脳先進的個別研究開発課題」、課題番号JP20dm0307031; 「革新的技術による脳機能ネットワークの全容解明プロジェクト」 JP21dm0207086), JST-PRESTO ("Pioneering Innovative Computing Technology", Project No. 19205833), and the Japan Science and Technology Agency (JST). Multicellular Biocomputing", project number 21H0516513; "Reconstruction of Humanities by Decoding Emotional Information", project number 21H05060).

Supplementary information and explanation of terms
Co-activation Pattern Analysis (CPA)
A method of analyzing brain activity. When one part of the brain is active, other parts of the brain are also active in various patterns. Co-activation pattern analysis extracts and visualizes such patterns from time-series data of brain activity.

2. Human Connectome Project
This is a leading open science project in the field of neuroscience, sponsored by the University of Washington and others. It measures a variety of brain activity data, including default mode, from over 1,000 volunteers and makes it available to researchers around the world.

PRESS RELEASE(PDF)