Diabetic retinopathy stage detection using convolutional fine-tuned transfer Learning model
Journal Title: International Journal of Experimental Research and Review - Year 2023, Vol 31, Issue 2
Abstract
Diabetic Retinopathy (DR) is a prevalent eye condition that occurs as a frequent complication among individuals with diabetes, particularly those who have been living with the disease for an extended period of time. This study uses fundus images to diagnose DR at five stages from early to late with No DR, Mild, Moderate, Severe, and Proliferative DR, commonly known as Stage 0 to Stage 4, respectively. This will aid in the timely treatment of diabetic patients preventing them from developing DR as early as possible. We used two most popular open-source datasets, the DR Detection database, namely APTOS 2019 and EyePACS, and combined them to create a larger dataset to trade off the data-centric obstacle and shortfall for any Deep Learning-based prediction models. Data augmentation and preprocessing techniques are applied to the images before feeding them to the proposed model to get a more accurate and efficient one. In the modern age oriented to Artificial Intelligence (AI), it is necessary to thoroughly analyze the identification of DR based on the existing Deep Learning (DL) models. After learning about the limitations of existing models, we have fine-tuned the ResNet50, DenseNet201 and InceptionV3 to enhance the model performance of the detection and categorization of DR. We have since proposed three Deep Convolutional Neural Networks (DCNN) models with better outcome based on accuracy than the existing state-of-the-art (SOTA) models. The fine-tuned DenseNet201 model, among the other two, performed significantly better with a validation accuracy of 90.04% and a negligible amount of loss, irrespective of each class, under the best configurable test conditions.
Authors and Affiliations
Jahnabi Medhi, Mithun Karmakar, Anup Kumar Barman, Subhash Mondal, Amitava Nag
Keywords
Related Articles
- EP ID EP719461
- DOI https://doi.org/10.52756/10.52756/ijerr.2023.v31spl.004
- Views 63
- Downloads 0
Evaluation of a Best Digital Supplier by Fuzzy SWARA-WASPAS Strategies
In this ruthless society, digital suppliers are noteworthy in building each organization to be productive and rich. Hence, choosing a reliable and well-grounded digital supplier becomes very necessary. The process of cho...
A Geo-Spatial analysis and assessment of groundwater potential zones by using remote sensing and GIS techniques-A micro level study of Bhagwanpur-I CD Block in Purba Medinipur District, West Bengal, India
The important resource of groundwater is a contributing significantly in total annual supply. However, over exploitation has depleted groundwater availability considerably and also led to land subsidence at some places....
Modulatory role of phyto-products in foraging behaviour of Drosophila sp leading to altered reproductive strategies: a novel experimental
Fruit flies are important in genetic purview but sometimes are nuisance to household commodity. Therefore, to restrict their uncontrolled trolling a preliminary experiment was conducted using several food items mixed wit...
Evaluation of a Probabilistic Framework for Traffic Volume Forecasting Using Deep Learning and Traditional Models
Forecasting is a very important issue in the present times to optimize urban traffic management systems and their proper planning. It is often found that traditional forecasting models cannot fully approximate the non-li...
Impact of backwardness on health-case study Pakhiralaya village, Gosaba Block, Sundarban, West Bengal, India
The Sundarban is a national park, biosphere reserve and tiger reserve in West Bengal, India. Pakhiralaya village is a part of Sundarban and it located exact opposite to the buffer area of Sundarban reserve forest. This v...