Khan, Sabiha and Reddy, Karuna G. (2023) Finding the determinants of lower limb amputations related to diabetic foot ulcer - a logistic regression classifier. In: PRICAI 2023: Trends in Artificial Intelligence. Lecture Notes in Computer Science . Springer Nature, Singapore, pp. 333-344. ISBN 978-981-99-7024-7
Full text not available from this repository.Abstract
Research is required for understanding the factors that contribute to Lower Extremity Amputation (LEA), especially in an area like Fiji that has a unique set of lifestyle factors and dietary habits. In such a developing country where T2DM-related amputations are one of the highest in the world, predicting the magnitude of the risk of LEA is vital for improving the care of Type 2 Diabetes Mellitus patients. Thus, this study developed a statistical model to predict quantifiable risk factors or predictors for LEA among T2DM patients from the three tertiary hospitals in Fiji. Such a model could possibly assist practitioners to understand the dynamics surrounding the problem and come up with possible solutions that may help reduce or prevent limb loss among diabetics. From the binary logistic regression classifier created via the 10-Fold Cross-Validation technique, we find that predictors such as length of stay (los), illness duration, the medical conditions of thrombocytosis and leukocytosis, gender, age category, hypertension, and low haemoglobin levels are key determinants of LEA. These predictors are statistically significant and have small to moderate effects on the outcome. The model has high sensitivity and performs very well, which indicates that it is correctly identifying a large portion of patients with amputations, thus minimizing the risk of false negatives.
| Item Type: | Book Chapter |
|---|---|
| Subjects: | H Social Sciences > HA Statistics |
| Divisions: | School of Information Technology, Engineering, Mathematics and Physics (STEMP) |
| Depositing User: | Karuna Reddy |
| Date Deposited: | 01 Feb 2024 22:58 |
| Last Modified: | 01 Feb 2024 22:58 |
| URI: | https://repository.usp.ac.fj/id/eprint/14415 |
Actions (login required)
 |
View Item |