Hybrid and transformer based artificial intelligence for diabetic retinopathy diagnosis a systematic review of methods challenges and clinical readiness
This systematic review examines 27 recent studies to map the evolving landscape of AI-driven DR diagnosis and argues that future efforts must prioritize lightweight, interpretable, and temporally aware AI systems that integrate multimodal patient data.