Related Experiment Video
Updated: Aug 6, 2026
05:17
Hybrid PET/MRI Imaging of Alzheimer's Disease Based on 18F-AV-1451
Published on: April 18, 2025
Feature integration of [18F]FDG PET brain imaging using deep learning for sensitive cognitive decline detection
Youjin Lee1, Seonguk Kim2, Sangil Kim1
1Department of Mathematics, Pusan National University, Busan, Republic of Korea.
Plos One
|July 21, 2026
Summary
This study developed a data-efficient framework integrating multi-scale PET imaging to improve cognitive decline detection. The combined approach significantly enhanced classification accuracy, offering a promising tool for early diagnosis.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Neuroscience
Background:
- Accurate diagnosis of cognitive decline (CD), including early Alzheimer's disease, is crucial for timely intervention.
- Positron emission tomography (PET) reveals metabolic brain changes in CD but faces limitations due to cost and radiation.
- A data-efficient framework is needed to maximize PET's clinical utility.
Purpose of the Study:
- To propose a novel framework integrating multi-scale PET representations (voxel-level and region-level) for improved CD detection.
- To enhance the clinical applicability of PET imaging in diagnosing cognitive impairment.
- To develop a data-efficient approach for machine learning-based CD classification.
Main Methods:
- Extracted voxel-level features using CNN and PCANet from [¹⁸F]FDG PET scans.
- Derived region-level features using DNN from standardized uptake value ratio measurements.
- Integrated voxel- and region-level features via direct concatenation for classification.
- Employed 5-fold cross-validation on 252 participants (118 cognitively normal, 134 with CD) from the Alzheimer's Disease Neuroimaging Initiative.
Main Results:
- The integrated DNN-CNN model achieved the highest classification accuracy (0.87 ± 0.05), outperforming single-representation models.
- This integrated approach showed a 6.33% improvement in accuracy and reduced standard deviation compared to DNN alone.
- Significant improvements were observed in Recall (0.77 to 0.88) and F1-Score (0.82 to 0.88).
- Predicted CD probability correlated significantly with Mini-Mental State Examination (MMSE) scores, with superior performance over MMSE-based classification.
Conclusions:
- Combining multi-scale PET representations with deep learning significantly enhances cognitive decline classification.
- The proposed framework demonstrates improved sensitivity and accuracy for detecting cognitive decline.
- Multi-scale FDG-PET representations show potential for machine learning-based diagnostic tools in neurology.