Identifying generalized anxiety disorder using resting state habenular circuitry.
Zijuan MaYuan ZhongChristina S HinesYun WuYuting LiManlong PangJian LiChiyue WangPeter T FoxNing ZhangChun WangPublished in: Brain imaging and behavior (2021)
Studies identify the habenula as a key subcortical component in anxiety, with a role in predicting error coding within the evaluative system. However, no clinical reports of generalized anxiety disorder (GAD) describe resting state functional connectivity of habenular circuits. We hypothesized that resting-state functional connectivities of habenula would show differences in neuroanatomical correlates of the evaluative system (prefrontal cortex, habenula) of patients with GAD. We obtained 22 patients with GAD and 21 HCs, matched for gender, age, and years of education. Resting-state functional connectivity of the habenula was assessed using a seed-based template imposed on whole brain MRI, which provided an objective and semi-automated segmentation algorithm in MNI space. Patients with GAD demonstrated enhanced connectivities in the bilateral premotor cortex, right ventrolateral prefrontal cortex, medial frontal cortex, as well as the left orbitofrontal cortex, and reduced connectivities in the left posterior cingulate cortex, and right pulvinar. Moreover, striking differences of abnormal connectivities between groups were observed via analysis of receiver operating characteristic curves (ROC) of statistically significant. These results including ROC curves suggest the potential importance of the habenula in evaluating and deciding to personally relevant reward-related information.
Keyphrases
- functional connectivity
- resting state
- prefrontal cortex
- deep learning
- machine learning
- healthcare
- magnetic resonance imaging
- high throughput
- quality improvement
- white matter
- contrast enhanced
- physical activity
- computed tomography
- convolutional neural network
- mass spectrometry
- diffusion weighted imaging
- magnetic resonance
- blood brain barrier
- molecularly imprinted
- single cell