An Extended Approach to Predict Retinopathy in Diabetic Patients Using the Genetic Algorithm and Fuzzy C-Means

Table 2

Quantitative comparative outcomes for segmentation of the retinopathy. This evaluation is conducted between our model and baseline studies. The assessments are based on Relative volume difference (RVD), Volume overlap error (VOE), and Dice similarity (DICE).

Method	Dice	RVD (%)	VOE (%)
Deep membrane [4]	92%	-2.46	6.16
Improved U-NET [6]	90%	3.74	5.78
PCNN model [2]	92%	3.55	6.41
Ant colony algorithm [73]	88%	-5.64	7.94
Pixel-based Segmentation [1]	88%	-4.91	7.42
Artificial Neural Network [74]	89%	5.37	8.12
Morphological Watershed [75]	87%	4.96	7.76
Presented FCM	86%	4.07	7.39
Presented FCM+GA	94%	2.32	4.28