Influence of Functional Magnetic Resonance Imaging Data Preprocessing Pipelines on the Accuracy of Schizophrenia...

Area of Science:

• Neuroimaging
• Computational Psychiatry
• Machine Learning

Background:

• Accurate classification of schizophrenia patients from healthy controls is crucial for diagnosis and treatment.
• Functional magnetic resonance imaging (fMRI) offers insights into brain function but requires extensive preprocessing.
• The choice of preprocessing pipeline can significantly influence machine learning model performance.

Purpose of the Study:

• To analyze the impact of various functional magnetic resonance imaging (fMRI) data preprocessing pipelines on classification accuracy.
• To compare the effectiveness of different preprocessing strategies for distinguishing schizophrenia patients from healthy controls.
• To provide recommendations for optimizing fMRI preprocessing pipelines for machine learning-based classification tasks.

Main Methods:

• Utilized fMRI data from 72 subjects.
• Applied seven distinct preprocessing pipelines to each dataset.
• Employed three feature vector construction algorithms (ReHo, FCM, FHR) and 15 machine learning classifiers (scikit-learn).

Main Results:

• No single preprocessing pipeline optimized performance across all feature extraction methods.
• Spatial smoothing enhanced accuracy for ReHo and FHR metrics but not FCM.
• Frequency filtering and median normalization substantially improved accuracy for ReHo and FCM metrics.

Conclusions:

• Recommended preprocessing pipelines tailored to specific feature extraction methods (ReHo, FCM, FHR).
• Highlighted the importance of spatial smoothing for FHR and filtering/normalization for ReHo and FCM.
• Advised caution with ICA filtering due to inconsistent effects on classification accuracy.