Explainable Transform Learning for Interpretable Medical Image Analysis
Bibliographic record
Abstract
The rapid adoption of deep learning methods in medical imaging has led to remarkable improvements in detection, segmentation, and diagnosis tasks. However, the widespread use of complex black-box models raises significant concerns regarding interpretability, transparency, and trustworthiness particularly in clinical decision-making where explainability is critical. This paper introduces a novel Explainable Transform Learning (ETL) framework designed to unify the strengths of data-driven representation learning with interpretable mathematical transforms. Unlike conventional convolutional or purely deep architectures, the proposed method explicitly integrates domain-relevant transform functions (e.g., wavelet, Fourier, and Radon transforms) into the learning pipeline, ensuring that intermediate features remain clinically interpretable. We demonstrate the framework’s effectiveness on multiple medical imaging modalities, including MRI, CT, and histopathology, where interpretability is often as important as accuracy. Experimental results show that ETL achieves competitive performance with state-of-the-art deep models while offering clear, human-understandable feature attributions that correlate with clinical markers. Moreover, our method reduces the dependency on large annotated datasets by leveraging structured priors, making it well-suited for data-scarce healthcare applications. This work contributes a step toward transparent and trustworthy artificial intelligence in medicine, highlighting that improved diagnostic performance need not come at the cost of explainability. We conclude with a discussion on how ETL can be extended for real-time clinical workflows and its potential integration into next-generation decision-support systems.
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 | 0.002 | 0.008 |
| 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 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 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".