Multimodal based classification of schizophrenia patients.

Area of Science:

• Neuroscience
• Psychiatry
• Medical Imaging

Background:

• Schizophrenia diagnosis relies on clinical assessments, prompting a search for objective, biologically based markers.
• Functional network connectivity is a promising area for predicting schizophrenia, but fMRI alone may limit insights into dysfunctional connectivity.

Purpose of the Study:

• To develop and validate a framework for classifying schizophrenia patients and healthy controls using combined fMRI and MEG data.
• To assess the efficacy of integrating functional network connectivity metrics from both fMRI and MEG for improved schizophrenia prediction.

Main Methods:

• Utilized resting-state fMRI and band-limited envelope correlation metrics from MEG to analyze functional network components.
• Developed a classification framework integrating data from both imaging modalities to differentiate schizophrenia patients from healthy controls.

Main Results:

• The combined fMRI and MEG approach significantly enhanced classification accuracy for schizophrenia.
• Accuracy improved by up to approximately 7% compared to fMRI alone and up to approximately 21% compared to MEG alone.

Conclusions:

• Combining fMRI and MEG provides valuable insights into brain network connectivity characteristics in schizophrenia.
• This integrated framework offers a more accurate and promising method for the prediction of schizophrenia.