Entropy-Guided Slice Selection for Weakly Supervised Binary ASPECTS Classification from Non-Contrast CT
Bibliographic record
Abstract
Timely and accurate assessment of ischemic damage is critical for acute stroke management. The Alberta Stroke Program Early CT Score (ASPECTS) is a standardized tool used to estimate infarct extent and guide treatment decisions. However, manual ASPECTS scoring on non-contrast computed tomography (NCCT) faces critical constraints: labor-intensive expert interpretation, substantial inter/intra-rater variability, time-consuming multi-region assessment, and poor sensitivity to early ischemic changes with subtle hypodensity. In this work, we propose a weakly supervised deep learning framework for binary ASPECTS classification ($\leq 6$vs.$>6$), the critical threshold for thrombectomy eligibility, to support treatment triage and prognosis without relying on pixel-level annotations. Our method begins with entropy-based slice selection to identify the most informative axial slices from each NCCT volume. A 3D convolutional neural network is then trained using volume-level binary ASPECTS labels, enhanced by auxiliary supervision from binary labels for each of the ten ASPECTS regions to encourage region-aware learning. This enables robust infarct pattern recognition while avoiding the need for detailed segmentation masks. Our method achieves strong binary ASPECTS classification performance (AUROC:$\mathbf{9 0. 3 6} \boldsymbol{\pm} \mathbf{8. 4 4}$) while substantially reducing annotation requirements, highlighting its promise for integration into realworld stroke triage workflows.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| 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 | 0.001 |
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".