Contextual Corpus Callosum Analysis for Differentiating Early and Late Mild Cognitive Impairment
Bibliographic record
Abstract
In recent times, deep learning extensively utilized across diverse domains, including Radiology. They play a pivotal role in early disease detection by analyzing data to identify patterns and biomarkers. Here, we propose a novel approach for classifying between Early MCI (EMCI) and Late MCI (LMCI), by integrating object detection with 3D CNN. Leveraging the capabilities of YOLOv5 object detection and 3D CNN, we focus on identifying changes in the corpus callosum (CC) and its neighbourhood, a crucial brain structure facilitating inter-hemispheric communication. Unlike existing methods that require manual cropping of CC, our approach integrates both detection and classification models, along with eliminating need for preprocessing steps. Applying on balanced dataset of Structural MRI volumes of 1098 subjects obtained from publicly available ADNI dataset, our method achieves accuracy of 88.41%. The novelty lies in integrating YOLOv5 object detection with 3D CNN for complete automation without the need for preprocessing, thereby utilizing "CC with context" for classification. The importance of utilizing neighbourhood of CC is illustrated by enhance classification accuracy of nearly 7%.Clinical relevance- This study demonstrates that the proposed 3D CNN model can accurately differentiate between early and late stages of Mild Cognitive Impairment by focusing on distinct subregions of the CC, aligning with known patterns of neurodegeneration. The model's interpretability to localize clinically relevant brain regions enhances its diagnostic value and supports more targeted, stage-specific interventions.
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) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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".