Blockchain technology is the trend in the 21st century with robust establishments in various societal contexts, including business, government, healthcare, and finance. Its diversified applications extend beyond cryptocurrencies and bitcoin, enabling data-sharing and control over access based on user type. This technology contrasts with machine learning, which focuses on data analysis and prediction. The additional advantage of using authentic datasets for prediction is the enhanced trust that can be gained from these predictions.Complete blindness can result from the chronic disease known as diabetic retinopathy. To lessen the likelihood of vision loss, the disease needs to be diagnosed early. Lesion segmentation was carried out on the image data after the data had been pre-processed using the median filtering technique. The Taylor African Vulture Optimization (AVO) algorithm was used to further refine the hyper-parameters in these data, and the most important characteristics were then fed into the SqueezeNet classifier to predict the development of diabetic retinopathy (DR) disease. The final output was saved in the blockchain architecture, where the EHR manager could access it and guarantee that only those with permission could access the prediction findings and associated patient data.