Adapting Prompt-Driven Approach for Diabetic Foot Ulcer Segmentation: A Sri Lankan Patient Study
Bibliographic record
Abstract
Diabetes is a common metabolic disorder affecting millions worldwide, including a significant population in Sri Lanka, and often leads to diabetic foot ulcers (DFUs). Accurate wound measurement is essential for diagnosing and monitoring DFUs, as it provides crucial metrics for tracking healing and guiding treatment decisions. However, manual measurement is time-consuming and prone to errors, potentially resulting in misdiagnoses and flawed records. Automating wound segmentation from medical images offers a promising solution, improving efficiency and patient outcomes. This study presents a prompt-driven segmentation approach to address the challenges of DFU segmentation, particularly in scenarios with limited data from Sri Lankan patients. By using user-provided bounding boxes, the model focuses on complex wound areas, improving accuracy in cases with unclear boundaries or sparse data. The research involves creating a diverse dataset of 300 samples from 180 patients at Dambulla Base Hospital, capturing variations in skin tone, ulcer size, shape, and surrounding tissue conditions. The proposed approach fine-tunes the Segment Anything Model (SAM) using prompt-driven techniques to adapt to the unique characteristics of DFUs in Sri Lanka. The model achieved strong performance metrics, including a Dice Coefficient of 83.31%, Specificity of 99.73%, and Precision of 90.42%. This method reduces the need for large annotated datasets, enhances gener-alization from limited data, and allows user-guided adjustments. By developing population-specific models, this research has the potential to improve clinical decision-making and significantly enhance DFU management in Sri Lanka. Furthermore, this approach demonstrates the broader applicability of prompt-driven segmentation techniques to address healthcare challenges in resource-constrained settings.
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) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 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".