993 Validating a Novel Machine Learning Tool for Objective Measurement of Clinical Burn Scar Assessment
Bibliographic record
Abstract
Abstract Introduction Hypertrophic scars are a common and challenging sequelae of burn injuries, often carrying functional and psychosocial impairment. Unfortunately, there lacks standardization in clinical burn scar assessment. Typical scoring systems such as the Patient and Observer Scar Assessment Scale (POSAS) are based on subjective observations which leaves room for bias and interrater variability. Alternative methods that measure objective features (e.g., size, thickness, colour, and elasticity) show promise, though they require tools that are costly and time prohibitive. Herein we evaluate the validity of a novel objective measurement tool that utilizes smartphone photography and machine learning (ML) to generate quantitative scar measures. Methods Burn patients recruited at a single centre underwent scar assessment using validated objective tools: ultrasound, pigmentation, and skin elasticity measurement devices. In tandem with the POSAS, we validated and trained our novel artificial intelligence-based tool with 2D colour images and video capture, with objective assessments as the best available gold standard. Results Ongoing collection and analyses on 60 burn scars are being conducted. Preliminary machine learning size and pigmentation analyses are at par with the device assessments. POSAS scores positively correlate with objective and ML model measures. Promising preliminary data demonstrates that the ML model can perform size and pigmentation analysis. Additional training and alterations to the algorithm are required to produce measurements of lesion thickness and elasticity. Conclusions This proof-of-concept study highlights the potential for ML models in burn scar assessment. Further development of this ML pipeline will allow for quick and accessible assessments of hypertrophic scars with mobile devices. Our approach could represent an alternative method for clinicians to track scar progression in patients with geographical barriers. Applicability of Research to Practice With the apparent lack of accessible objective assessments for burn scars, this project addressed the barrier to provide clinicians with concrete measures. Accurately tracking the progression of burn scars in response to therapies can dictate patients care plan. As there is a widespread use of smart device photography in the charting of burn scars, incorporating ML technology proves feasible to the workflow. Once fully developed, loading our algorithm within an app or tool on smart devices will make objective burn scar assessments available and consistent across different sites. Funding for the Study N/A
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 | 0.004 | 0.011 |
| 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 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
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".