
Research Article
Use of Raman Spectroscopy to Diagnose Diabetes with SVM
@INPROCEEDINGS{10.1007/978-3-031-28790-9_6, author={Le Anh Duc and Nguyen Thanh Tung}, title={Use of Raman Spectroscopy to Diagnose Diabetes with SVM}, proceedings={Nature of Computation and Communication. 8th EAI International Conference, ICTCC 2022, Vinh Long, Vietnam, October 27-28, 2022, Proceedings}, proceedings_a={ICTCC}, year={2023}, month={3}, keywords={Diabetes Raman spectroscopy Machine learning SVM Diagnose}, doi={10.1007/978-3-031-28790-9_6} }
- Le Anh Duc
Nguyen Thanh Tung
Year: 2023
Use of Raman Spectroscopy to Diagnose Diabetes with SVM
ICTCC
Springer
DOI: 10.1007/978-3-031-28790-9_6
Abstract
In this work, investigations were made for exploring the potential of machine learning in predicting Type 2 Diabetes Mellitus patients. In overall, 20 patients were assessed. The Raman spectrum was observed in four anatomical locations of the body: ear lobe, inner arm, thumb nail and cubital vein. The measurements were taken to examine the difference between Control and DM2 (9 well-controlled patients and 11 diabetic patients). To create effective diagnostic algorithms for categorization among these categories, multivariate approaches such as principal component analysis (PCA) paired with support vector machine (SVM) were applied. Based on the implemented classification systems, diabetic patients are classified using PCA-SVM shows the best potential of 80% in accuracy. Therefore, the taken approach successfully creates a classification and evaluation system. Overall, our findings show that the combination of Raman spectroscopy, PCA-SVM has several advantages in terms of preciseness and is suggested as a viable non-invasive diagnostic technique for diabetes.