Diabetic Foot Risk Classification using Decision Tree and Bio-Inspired Evolutionary Algorithms
B G Sudha1, V Umadevi2, Joshi Manisha Shivaram3, Mohamed Yacin Sikkandar4, Belehalli Pavan5, Abdullah Al Amoudi6
1B G Sudha, Department of CSE, B.M.S. College of Engineering, Bangalore (Karnataka), India.
2V Umadevi, Department of CSE, B.M.S. College of Engineering, Bangalore (Karnataka), India.
3Joshi Manisha Shivaram, Department of Medical Electronics, B.M.S. College of Engineering, Bangalore (Karnataka), India.
4Mohamed Yacin Sikkandar, College of Applied Medical Sciences, Majmaah University, Kingdom of Saudi Arabia.
5Belehalli Pavan, Podiatry, Karnataka Institute of Endocrinology and Research, Bangalore (Karnataka), India.
6Abdullah Alamoudi, College of Applied Medical Sciences, Majmaah University, Kingdom of Saudi Arabia.
Manuscript received on 04 December 2019 | Revised Manuscript received on 12 December 2019 | Manuscript Published on 31 December 2019 | PP: 232-239 | Volume-9 Issue-2S December 2019 | Retrieval Number: B10811292S19/2019©BEIESP | DOI: 10.35940/ijitee.B1081.1292S19
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Diabetic foot complications are a burden to the Indian population which affects both financially and physically. The complications could be prevented if the risk of diabetic foot are detected well in advance before the peripheral nerves are damaged leading to amputation and limb loss. The quantification of severity plays an important role in timely intervention, delivery of appropriate treatment and prevention of amputation. This can be modeled as a classification problem where the risk category is stratified into different levels of severity. This paper is an approach to build such a system, capable of classifying the risk category of diabetic patients for suitable follow-up and care. Decision trees are used for the same with features selected using bio-inspired evolutionary algorithms like Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Cuckoo Search (CS), FireFly (FF), Dragon Fly (DF) and Gravitational Search Algorithm (GSA). The overall accuracy is 77% but it identifies the low risk and high risk cases effectively with 97% and 89% respectively.
Keywords: Diabetic Foot Risk Classification, Decision Tree, Feature Selection, Bio-Inspired Algorithms.
Scope of the Article: Classification