Diagnostic Markers of Subclinical Depression Based on Functional Connectivity

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Major depression disorder is often accompanied by the dysfunction of the lateral habenula nucleus (LHb). The onset of subclinical depression is often a precursor to the onset of major depression disorder (MDD), and the brain function patterns of patients with subclinical depression are similar to those of patients with major depression. Therefore, it seems that the functional connection pattern of LHb and its surrounding regions can be used to predict subclinical depression. This chapter reviews a subclinical depression prediction model that detects abnormal functional connections through machine learning. By means of resting-state functional magnetic resonance imaging (fMRI), the functional connections of the brain regions were calculated, and the functional brain network was modeled. Functional connections with significant differences were entered into the machine learning model as classifier features. Anomalous functional connections, which have a greater impact on classification results, are evaluated for accuracy, receiver operating characteristic curve (ROC), and area under curve (AUC). Based on the analysis of the functional connection network of the whole brain, we found five brain regions having the greatest difference between the subclinical depression (SD) group and the healthy control (HC) group: anterior cingulate cortex, thalamus, superior parietal lobule, superior frontal gyrus, and posterior superior temporal sulcus. Among the thalamus-related functional connections, the right posterior parietal thalamus, where LHb located, has the most abnormal functional connections, the number of which is 18, and in the left LHb, there are only 3 abnormal functional connections. At a significant level p < 0.01, the network node degree of ten subregions of the thalamus in the SD group showed a significant difference from the corresponding node degree in the HC group. Therefore, we believe that functional connection analysis based on the thalamus and habenula nucleus can provide a high accuracy biomarker for the prediction of subclinical depression.

Original languageEnglish (US)
Title of host publicationContemporary Clinical Neuroscience
PublisherSpringer Nature
Pages283-296
Number of pages14
DOIs
StatePublished - 2021

Publication series

NameContemporary Clinical Neuroscience
ISSN (Print)2627-535X
ISSN (Electronic)2627-5341

Keywords

  • Brain biomarker
  • Brain network
  • Functional connection
  • Machine learning
  • Node degree
  • Resting-state functional MRI
  • Subclinical depression

ASJC Scopus subject areas

  • Behavioral Neuroscience
  • Cognitive Neuroscience
  • Neurology
  • Sensory Systems
  • Clinical Neurology

Fingerprint

Dive into the research topics of 'Diagnostic Markers of Subclinical Depression Based on Functional Connectivity'. Together they form a unique fingerprint.

Cite this