Automated detection of diabetic retinopathy in fundus images using fused features

Early and precise detection of diabetic retinopathy prevents vision impairments through computer-aided clinical procedures. Identifying the symptoms and processing those by using sophisticated clinical procedures reduces hemorrhage kind of risks. The input diabetic retinopathy images are influenced by using computer vision-based processes for segmentation and classification through feature extractions. In this article, a delimiting segmentation using knowledge learning (DS-KL) is introduced for classifying and detecting exudate regions by using varying histograms. The input image is identified for its histogram changes from the feature-dependent segmentation process. Depending on the training knowledge from multiple inputs with different exudate regions, the segmentation is performed. This segmentation identifies infected and noninfected regions across the delimiting pixel boundaries. The knowledge-learning process stores the newly identified exudate region for training and pixel correlation. The recurrent training improves the segmentation accuracy with precise detection and limited errors.

show abstract

“…e methods AD-FF [12] and PSO-GIT2FMFs [11] are considered alongside the proposed DS-KL method in the metrics comparison.…”

Section: Implementation and Discussionmentioning

confidence: 99%

“…e proposed method achieves a high accuracy rate in the detection process, reducing the error rate in the diagnosis process. Bibi et al [12] proposed an automated DR detection method based on fundus images. Symptoms and microaneurysms are identified based on fundus images that capture the actual information about DR.…”

Section: Related Workmentioning

confidence: 99%

Pixel-Boundary-Dependent Segmentation Method for Early Detection of Diabetic Retinopathy

Sandhya

Suhasini

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…After the segmentation operation, various properties including shape, color, and size with exudates (hard and soft), hemorrhages, and microaneurysms have appeared. The circular shape and the dark red color with small in size is the nature of MAs 32 . By using deep LSSTM model, the features are automatically extracted or further processed.…”

Section: Proposed Methodologymentioning

confidence: 99%

Deep long and short term memory based Red Fox optimization algorithm for diabetic retinopathy detection and classification

Priya

Sivarani

Saravanan

2021

Numer Methods Biomed Eng

View full text Add to dashboard Cite

Because of retina abnormalities of diabetic patients, the most common vision‐threatening disease is diabetic retinopathy (DR). The DR diagnosis and prevention are challenging tasks as they may lead to vision loss. According to the literature analysis, the shortcomings in existing studies, such as failed to reduce the feature dimension, higher execution time, and higher computational cost, unable to tune the hyper‐parameters, such as a number of hidden layers and learning rate, more computational complexities, higher cost, and so forth, during DR classification. To tackle these problems, we proposed a deep long‐ and short‐term memory (LSTM) in a neural network with Red Fox optimization (deep LSTM‐RFO) algorithm for DR classification. The four major components involved in the proposed methods are image preprocessing, segmentation, feature extraction, and classification. At first, an adaptive histogram equalization and histogram equalization model performs the fundus image preprocessing, thereby neglecting the noise and improving the contrast level of an image. Next, an adaptive watershed segmentation model effectively segments the lesion region based on the optic disc color and size of hemorrhages. At the third stage, we have extracted statistical, intensity, color, and shape features. Finally, the single normal class with three abnormal classes such as mild non‐proliferative diabetic retinopathy, moderate NPDR, and severe NPDR are accurately classified using the deep LSTM‐RFO algorithm. Experimentally, the MESSIDOR, STARE, and DRIVE datasets are used for both training and validation. MATLAB software performs the implementation process with respect to various evaluation criteria used. However, the proposed method accomplished superior performance, such as 98.45% specificity, 96.78% sensitivity, 97.92% precision, 96.89% recall, and 97.93% F‐score results in terms of DR classification than previous methods.

show abstract

“…Hence, early identification of DR is very important and is made possible by ophthalmoscopic investigations during regular physical eye examinations. During an ophthalmoscopic examination (also known as a funduscopy) an ophthalmologist looks into the interior surface of the eye using an ophthalmoscope to identify pathognomonic indications of DR such as hemorrhages (HMs), soft/hard exudates (EXs), and microaneurysms (MAs) [16] [7].…”

Section: Introductionmentioning

confidence: 99%

“…Recent advances in fundus photography [6] allow CAD systems to detect and identify the structure of interest in a semi-automatic or fully automated manner [13] [14], making it easy to distinguish between normal and DR fundus images. Therefore, the CAD systems can be used during the pre-screening phases to address the following issues, like limited accessibility of ophthalmic medical professionals and an increasing number of DR patients, intrinsic variability in diagnostic actions and quality across communities, substantial social and personal burden of blindness and low vision [7]. Moreover, Deep learning techniques show excellent performance in identifying DRs and offer a high level of accuracy that sets them apart from other models.…”

Section: Introductionmentioning

confidence: 99%

EHSCO-Deep Residual Network: Elephant Herding Sine Cosine Optimization-based Deep Residual network for Diabetic Retinopathy Detection

Gandle¹

2022

JNACS

View full text Add to dashboard Cite

Diabetic retinopathy (DR) is a widespread problem of diabetes and is one of the main causes of loss of sight in the dynamic population. Several complications of DR can be prevented by timely behavior and blood glucose control. In this paper, the Elephant Herding Sine Cosine Optimization-based Deep Residual network (EHSCO-based Deep Residual Network) is developed for the DR detection. Here, pre-processing is performed for the input image using the grayscale conversion and Type 2 Fuzzy and Cuckoo Search-Based Filter (T2FCS). After that, the optic disc detection is carried out using Sparse Fuzzy C-Means clustering (Sparse FCM), whereas the blood vessel detection is performed using Morphological Top-Hat transform for the pre-processed input image. Besides, feature extraction is performed to extract the features, such as Texton, Walsh Hadamard transform, mean, variance, entropy and kurtosis. Finally, the extracted features are given to the detection module, which is carried out based on Deep Residual Network classifier. In addition, the employed classifier is trained by the developed EHSCO optimization algorithm. Here, the developed EHSCO-based Deep Residual Network is the combination of Elephant Herding Optimization (EHO) and Sine Cosine Algorithm (SCA). Besides, the developed DR detection technique obtained efficient performance in terms of maximum True Positive Rate (TPR) with 0.963, maximal True Negative Rate (TNR) with 0.959, and higher accuracy with 0.956 with respect to DIARETDB0 dataset respectively.

show abstract

Automated detection of diabetic retinopathy in fundus images using fused features

Cited by 20 publications

References 32 publications

Pixel-Boundary-Dependent Segmentation Method for Early Detection of Diabetic Retinopathy

Pixel-Boundary-Dependent Segmentation Method for Early Detection of Diabetic Retinopathy

Deep long and short term memory based Red Fox optimization algorithm for diabetic retinopathy detection and classification

EHSCO-Deep Residual Network: Elephant Herding Sine Cosine Optimization-based Deep Residual network for Diabetic Retinopathy Detection

Contact Info

Product

Resources

About