Health System for Exercise Rehabilitation Detection Retina Images and IoT Blockchain
DOI:
https://doi.org/10.15379/ijmst.v10i1.2987Keywords:
Medical Diagnostic, Retina Images, Internet of Things, Diabetic Retinopathy, CNNAbstract
Smart Healthcare which is based on deep learning is becoming increasingly popular because of its practical applications and has grown popular after its incorporation with IoT. Degenerative eye disease is the main factor of blindness in people of working age. Asian nations with large populations, like India and China, are on the edge of a diabetes epidemic. In terms of medical screening and diagnosis, a large number of diabetes patients posed a huge problem for skilled clinicians. The idea is to employ deep learning algorithms to detect blind spots in the eye and estimate the severity of the stage. We present an optimum approach for detecting blindness in retinal images based on recently released pre-trained EfficientNet models, as well as a comparative assessment of many innovative neural network models, in this study. On a benchmark dataset of retina images obtained by diagnostic imaging at various imaging phases, our EfficientNet-B5-based model assessment performs better than CNN and ResNet50 models.