Detection and Comparison of Diabetic Retinopathy using Thresholding Algorithm and CMeans Clustering Algorithm
DOI:
https://doi.org/10.47750/pnr.2022.13.S04.021Keywords:
Novel Diabetic Retinopathy Detection, Machine learning, Thresholding Algorithm, C-Means Clustering Algorithm, MATLAB Programming, Peak Signal to Noise Ratio (PSNR).Abstract
Aim: The aim of this research work is to detect the presence of Novel Diabetic Retinopathy Detection using modern algorithms, and comparing the peak signal to noise ratio (PSNR) between Thresholding Algorithms and C-Means Clustering Algorithms.
Materials and Methods: The sample images were taken from kaggle’s website. Samples were considered as (N=24) for Thresholding Algorithm and (N=24) for c-means clustering algorithm in accordance with total sample size calculated using clinicalc.com by keeping alpha error-threshold value 0.05, enrollment ratio as 0.1, 95% confidence interval, G power as 80%. The PSNR was calculated by using the novel MATLAB Programming with a standard data set.
Results: Comparison of PSNR is done by independent sample t-test using SPSS software. There is a statistical significant difference between Thresholding Algorithm and C-means clustering algorithm with p=0.014, p<0.05 (PSNR = 37.290) showed better results in comparison to Thresholding Algorithm (PSNR =14.7327).
Conclusion: C-Means Clustering Algorithms were found to give higher PSNR than in Thresholding Algorithms for the Novel Diabetic Retinopathy Detection.
