A novel convolutional interpretability model for pixel-level interpretation of medical image classification through fusion of machine learning and fuzzy logic
Bibliographic record
Abstract
Artificial intelligence (AI) models for medical image analysis have achieved high diagnostic performance, but they often lack interpretability, limiting their clinical adoption. Existing methods can explain predictions at the image level, but they cannot provide pixel-level insights. This study proposes a novel fusion of machine learning and fuzzy logic to develop an interpretable model that can precisely identify discriminative image regions driving diagnostic decisions and generate heatmap visualization. The model is trained and evaluated on a dataset of CT scans containing healthy and diseased organ images. Quantitative features are extracted across pixels and normalized into representation matrices using a machine learning model. Subsequently, the contribution of each detected lesion to the overall prediction is quantified using fuzzy logic. Organ segment weighted averages are computed to identify significant lesions. The model explains application of AI in medical imaging with an unprecedented level of detail. It can explain fine-grained image areas that have the greatest influence on diagnostic outcomes by mapping raw image pixels to fuzzy membership concepts. Lesions are found with effect sizes and statistical significance (p < 0.05). Our model outperforms three existing methods in terms of interpretability and diagnostic accuracy by 10–15%, while maintaining computational efficiency. By disclosing crucial image evidence that supports AI decisions, this interpretable model improves transparency and clinician trust. Ethical implications of integrating AI in clinical settings are discussed, and future research directions are outlined. This study significantly advances the development of safe and interpretable AI for enhancing patient care through imaging analytics.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 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 | 0.000 |
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".