Plantar Thermogram Analysis Using Deep Learning for Diabetic Foot Risk Classification
Authors' conclusion
These results suggest that thermography combined with deep learning could be developed for screening purposes to stratify patients at risk of developing diabetic foot ulcers.
Study context
- Population
- Adults with diabetes undergoing diabetic foot screening; mean age 63.1 ± 12.6 years, 52.3% female, 62.1% with diabetes >10 years, mean BMI 27.5 ± 5.6 kg/m²
- Participants
- 153 thermal images
- Outcome
- Classification of diabetic foot ulcer risk (IWGDF category 0 vs 1–3)
Methodological detailsprospective, deep learning, five-fold cross-validation, Trained on balanced dataset of 98 images (49 normal, 49 abnormal; 80% training, 20% validation); tested on separate dataset of 55 images (42 normal, 13 abnormal)
- Prospective
- Deep learning
- Five-fold cross-validation
- Trained on balanced dataset of 98 images (49 normal, 49 abnormal; 80% training, 20% validation); tested on separate dataset of 55 images (42 normal, 13 abnormal)
Performance
As reported in the abstract
| Test or model | Accuracy | Sensitivity | Specificity | Matthews correlation coefficient |
|---|---|---|---|---|
| Neural network model | 71.8 ± 4.9% | 81.2 ± 10.0% | 64.0 ± 7.4% | 0.46 ± 0.08% |
Neural network model
Accuracy71.8 ± 4.9%
Sensitivity81.2 ± 10.0%
Specificity64.0 ± 7.4%
Matthews correlation coefficient0.46 ± 0.08%
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Plantar Thermogram Analysis Using Deep Learning for Diabetic Foot Risk Classification. J Diabetes Sci Technol. Published 20 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →