3D CNN Based Automatic Diagnosis of Attention Deficit Hyperactivity Disorder Using Functional and Structural MRI
Bibliographic record
Abstract
Attention deficit hyperactivity disorder (ADHD) is one of the most common mental health disorders. As a neuro development disorder, neuroimaging technologies, such as magnetic resonance imaging (MRI), coupled with machine learning algorithms, are being increasingly explored as biomarkers in ADHD. Among various machine learning methods, deep learning has demonstrated excellent performance on many imaging tasks. With the availability of publically-available, large neuroimaging data sets for training purposes, deep learning-based automatic diagnosis of psychiatric disorders can become feasible. In this paper, we develop a deep learning-based ADHD classification method via 3-D convolutional neural networks (CNNs) applied to MRI scans. Since deep neural networks may utilize millions of parameters, even the large number of MRI samples in pooled data sets is still relatively limited if one is to learn discriminative features from the raw data. Instead, here we propose to first extract meaningful 3-D low-level features from functional MRI (fMRI) and structural MRI (sMRI) data. Furthermore, inspired by radiologists' typical approach for examining brain images, we design a 3-D CNN model to investigate the local spatial patterns of MRI features. Finally, we discover that brain functional and structural information are complementary, and design a multi-modality CNN architecture to combine fMRI and sMRI features. Evaluations on the hold-out testing data of the ADHD-200 global competition shows that the proposed multi-modality 3-D CNN approach achieves the state-of-the-art accuracy of 69.15% and outperforms reported classifiers in the literature, even with fewer training samples. We suggest that multi-modality classification will be a promising direction to find potential neuroimaging biomarkers of neuro development disorders.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".