Chinese scholars have made progress in multimodal datasets and related research for mental disorders

Figure Data-driven quantitative research methods

According to WHO statistics and forecasts, there are more than 300 million people with depression in the world, and depression will rank first
in the global burden of disease by 2030.
China has clearly pointed out in the "Healthy China 2030" planning outline: "It is necessary to strengthen the intervention of common mental disorders and psychological and behavioral problems such as depression and anxiety"
.
At present, the diagnosis method of mental disorders such as depression is usually a subjective interview based on a scale, which is easily affected by environmental and individual differences, resulting in a clinical recognition rate of only 47.
3% of depressive disorders, so convenient and objective and effective quantitative evaluation methods
are urgently needed.
With the rapid development of artificial intelligence technology, the use of physiological signals to objectively, quickly and efficiently assess mental disorders such as depression is conducive to the early diagnosis and treatment
of mental disorders.
With the support of the National Natural Science Foundation of China (grant numbers: 61632014, 62076113, 62072219, 61210010, 60973138), Professor Hu Bin's team of Lanzhou University has made progress
in multimodal datasets and related research for mental disorders.

The research team has long been engaged in the research of quantitative perceptual computing of mental disorders based on universal physiological and behavioral data, and proposed the concept of "psychophysiological computing" in 2012, which established a data-driven quantitative research method
for cognitive function and psychological state by integrating cognitive science, psychology and information science 。 Aiming at the diversity, temporal variability and individual differences of physiological and behavioral phenotypes, this method focuses on the complex interaction between brain-body-psychology, and quantifies the perception of human advanced cognitive functions (such as emotions) by constructing a multi-level spatiotemporal aggregation computing framework for multimodal psychophysiological data, which provides a new way
to break through the key technology of "sense" + "knowledge" of psychological and physiological information and realize the objective quantitative assessment of mental disorders in natural situations 。 In 2019, the team released MODMA, the world's first multimodal psychophysiological database for quantitative analysis of mental disorders, and the relevant research results were published in Scientific Data (Nature) in April 2022 under the title "A multi-modal open dataset for mental-disorder analysis"
。 Article link: _istranslated="1">.

Participants in the MODMA dataset included patients with clinical depression (diagnosed and selected by a psychiatrist) and normal population controls
.
The dataset includes three parts
: EEG experimental data of 128 electrodes of full conduction, wearable 3-lead EEG experimental data, and speech experimental data.
Among them, it included 128 induced EEG signals recorded at rest and during the Dot probe task (53 participants); 3-lead EEG signals recorded at rest (55 participants); Audio signals recorded while interviewing, reading, and viewing pictures (52 participants).

The dataset currently receives 566 applications
from more than 100 scientific institutions around the world.
In addition, the team held the MODMA Data Analysis Competition at the IEEE Healthcom 2020 International Conference, with a total of 266 teams from all over the world participating
.
The MODAM dataset and related work provide new ideas and methods for the diagnosis and treatment of mental disorders, actively promote the in-depth research in this field, and further expand the international academic influence
.
Dataset link: http://modma.
lzu.
edu.
cn/data/index/
.