Machine Learning for Dementia Prediction: A Systematic Review and Future Research Directions

Area of Science:

• Artificial Intelligence and Machine Learning in Healthcare
• Neurological Disease Prediction
• Biomedical Data Analysis

Background:

• Healthcare extensively utilizes Artificial Intelligence (AI) and Machine Learning (ML) for automated solutions.
• Dementia detection using ML methods is a significant research focus for automated disease prediction.
• Existing systematic literature reviews (SLR) on ML for dementia detection often focus on single data modalities.

Approach:

• This study conducts a comprehensive evaluation of ML-based automated diagnostic systems for dementia.
• It analyzes research articles from 2011-2022 focusing on dementia, machine learning, feature selection, data modalities, and automated diagnostic systems.
• The review critically analyzes and discusses selected articles, considering diverse data modalities like images, clinical features, and voice data.

Key Points:

• Machine learning models utilizing image data demonstrate superior performance in dementia prediction compared to clinical features or voice data.
• The study systematically reviews various data modalities for ML-based dementia detection.
• Identified limitations in current automated dementia prediction methods.

Conclusions:

• Image data-driven ML models offer a promising avenue for early dementia prediction.
• Further research is needed to overcome limitations in existing automated dementia detection systems.
• This SLR provides insights into future directions for improving ML-based dementia diagnosis.