The phobic brain: Morphometric features correctly classify individuals with small animal phobia

Area of Science:

• Neuroscience
• Psychiatry
• Machine Learning

Background:

• Specific phobia is an anxiety disorder involving intense fear of specific stimuli.
• Small animal phobia (SAP) is understudied, with previous research limited by small sample sizes and univariate analyses.
• Neural underpinnings of SAP require further investigation using advanced methodologies.

Purpose of the Study:

• To develop a machine learning model for classifying individuals with SAP based on gray matter features.
• To assess the contribution of structural brain networks in differentiating SAP patients from healthy controls.
• To elucidate the neurobiological basis of small animal phobia.

Main Methods:

• Employed a binary support vector machine (SVM) learning model.
• Analyzed gray matter features in 32 subjects with SAP and 90 matched healthy controls.
• Investigated whole-brain structures and specific macro-networks (Default Mode, Salience, Executive, Affective).

Main Results:

• A significant predictive model identified brain regions involved in emotional regulation, cognitive control, and sensory integration (cerebellum, temporal pole, frontal cortex, amygdala, thalamus).
• The Default Mode, Affective, and Central Executive networks significantly contributed to classifying individuals with SAP.
• Machine learning successfully differentiated SAP subjects from controls based on structural brain data.

Conclusions:

• This study provides novel insights into the neural correlates of small animal phobia.
• The findings highlight the potential of machine learning in understanding anxiety disorders.
• Results suggest new avenues for research and potential diagnostic strategies for SAP.