Ensemble learning using traditional machine learning and deep neural network for diagnosis of Alzheimer’s disease
Bibliographic record
Abstract
In recent years, Alzheimer's disease (AD) diagnosis using neuroimaging and deep learning has drawn great research attention. However, due to the scarcity of training neuroimaging data, many deep learning models have suffered from severe overfitting. In this study, we propose an ensemble learning framework that combines deep learning and machine learning. The deep learning model was based on a 3D-ResNet to exploit 3D structural features of neuroimaging data. Meanwhile, Extreme Gradient Boosting (XGBoost) machine learning was applied on a voxel-wise basis to draw the most significant voxel groups out of the image. The 3D-ResNet and XGBoost predictions were combined with patient demographics and cognitive test scores (Mini-Mental State Examination (MMSE) and Clinical Dementia Rating (CDR)) to give a final diagnosis prediction. Our proposed method was trained and validated on brain MRI brain images of the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. During the training phase, multiple data augmentation methods were employed to tackle overfitting. Our test set contained only baseline scans, i.e., the first visit scans since we aimed to investigate the ability of our approach in detecting AD during the first visit of AD patients. Our 5-fold cross-validation implementation achieved an average AUC of 100% during training and 96% during testing. Using the same computer, our method was much faster in scoring a prediction, approximately 10 min, than feature extraction-based machine learning methods, which often take many hours to score a prediction. To make the prediction explainable, we visualized the brain MRI image regions that primarily affected the 3D-ResNet model's prediction via heatmap. Lastly, we observed that proper generation of test sets was critical to avoiding the data leakage issue and ensuring the validity of results.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".