An explainable machine learning approach for Alzheimer’s disease classification
importance 8/10
ai assisted
Dementia and Cognitive Impairment Research Alzheimer's diagnostics stage: research
Problem
Early Alzheimer's diagnosis is difficult because subtle biomarker changes are overlooked and data is high-dimensional with limited samples.
Why it matters
Early identification of at-risk individuals can enable earlier treatment decisions.
What the AI did
An explainable ML pipeline reduced 1024 features and trained SVM models on 169,408 NACC records, tested on an external dataset.
Results
F1 98.9% binary (NC vs AD) and 90.7% multiclass on an external dataset.
Impact
External-dataset testing is stated, supporting the reported classification performance; clinical adoption not stated.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Abbas Saad Alatrany, Wasiq Khan, Abir Jaafar Hussain, Hoshang Kolivand, Dhiya Al‐Jumeily |
| Citations | 167 (OpenAlex, at import time) |
| References | 55 |
| Model or method | Support vector machine (SVM) |
| Organization | National Alzheimer's Coordinating Center (data source) |
| Primary source | https://doi.org/10.1038/s41598-024-51985-w |
| Open access | gold |
| Keywords | Alzheimer's disease, machine learning, explainability, SVM, classification, biomarkers |
Importance rating
8
major — millions of lives, an industry or a field
Early Alzheimer's diagnosis is a significant challenge per abstract; early identification enables earlier treatment decisions, but no population scale is stated.