A survey of deep learning techniques in detecting neurological disorders using MRI

Area of Science:

• Neuroimaging and Artificial Intelligence
• Medical Image Analysis
• Neurological Disorder Diagnostics

Background:

• Manual interpretation of MRI scans for neurological disorders is time-consuming and variable.
• Increasing global burden of neurological diseases necessitates automated diagnostic tools.
• Deep learning (DL) shows promise for automated feature extraction in neuroimaging.

Purpose of the Study:

• To provide a comprehensive analysis of deep learning approaches for MRI-based detection of neurological disorders.
• To systematically review recent advancements (2019-2025) in DL for neuroimaging.
• To identify trends, limitations, and future directions in the field.

Main Methods:

• Systematic review of 47 research articles (2019-2025).
• Analysis of over 40 deep learning architectures on 34 datasets.
• Categorization and critical examination of CNNs, ViTs, and hybrid models.

Main Results:

• Comparative analysis of DL model performance across various neurological conditions.
• Evaluation of dataset characteristics, protocols, and computational demands.
• Identification of emerging architectural trends and existing approach limitations.

Conclusions:

• Deep learning models show potential for accurate and scalable MRI-based neurological disorder detection.
• Gaps exist in validation, dataset heterogeneity, explainability, and clinical deployment.
• Future research should focus on multimodal learning and self-supervised methods for enhanced clinical translation.