Table 12. Comparative study of the proposed approach with previous work.

Reference	Year	No. of Classes	Model	Results
[28]	2021	29	Ensemble CNN	F1 Score 94.32%
[29]	2022	29	Ensemble Learning	Sensitivity for all 29 Classes Ranging from 0.00-1.00
[31]	2022	4	Deep Learning	RVO 88.4%
				DR 85.2%
				CSR 93.8%
				Healthy 86.2%
[12]	2023	4	EyeDeep-Net	Accuracy:
				Validation 82.13%
				Testing 76.04%
[27]	2023	4	Convolutional Ensemble	Accuracy 79.2%
[32]	2024	5	Deep Learning	Accuracy 89.10%
[34]	2024	2	ResNet152	ResNet152 89.17%‘
			Vision Transformer	Transformer 87.26%
			InceptionResNetV2	Inc.ResNetV2 88.11%
			RegNet	RegNet 88.54%
			ConVNext	ConVNext 89.08%
Proposed	2024	4	CNN	MH 93.17%
				Overall Acc. 89.81%
				DR 91.95%,
				ODC 94.60%
				WNL 92.43%