Disentangling shared and unique brain functional changes associated with clinical severity and cognitive phenotypes...

Area of Science:

• Neuroscience
• Artificial Intelligence
• Psychiatry

Background:

• Schizophrenia is characterized by cognitive impairments and altered brain function.
• Understanding the neural basis of cognitive deficits and symptom severity in schizophrenia is crucial.
• Shared and unique brain functional patterns remain poorly understood.

Purpose of the Study:

• To develop an interpretable graph-based multi-task deep learning framework.
• To simultaneously predict schizophrenia illness severity and cognitive functioning.
• To identify shared and unique brain functional patterns associated with these phenotypes.

Main Methods:

• Utilized functional connectivity data from 378 subjects across three datasets.
• Employed a novel graph-based multi-task deep learning framework.
• Validated performance against single-task and state-of-the-art multi-task learning methods.
• Confirmed findings using meta-analysis at regional and modular levels.

Main Results:

• The proposed framework significantly outperformed existing methods in predicting PANSS subscales and cognitive domain scores.
• Performance was consistent and replicable across three independent datasets.
• Identified both shared and unique functional brain patterns related to illness severity and cognitive deficits.

Conclusions:

• The study provides novel insights into the neural correlates of schizophrenia illness severity and cognitive impairments.
• The identified brain patterns offer potential targets for treatment evaluation and longitudinal studies.
• The interpretable deep learning framework demonstrates efficacy in psychiatric neuroimaging research.