%0 Journal Article %1 sanya2021novel %A Sanya, Waheed %A Bajpai, Gaurav %A Amour, Maryam %A Alawi, Mahmoud %A Adnan, Ali %D 2021 %J International Journal of Image Processing (IJIP) %K CLAHE, Diabetic Disc, Image, Mathematical Microaneurysms Morphological Operation, Optic Retina Retinopathy, %N 3 %P 22-36 %T A Novel Advanced Approach Using Morphological Image Processing Technique for Early Detection of Diabetes Retinopathy %U https://www.cscjournals.org/library/manuscriptinfo.php?mc=IJIP-1218 %V 15 %X Diabetic retinopathy (DR) is a common complication of diabetes mellitus and can lead to irreversible blindness. To date, DR is the leading cause of blindness and visual impairment among working adults globally. However, this blindness can be prevented if DR is detected early. Diabetes mellitus slowly affects the retina by damaging retinal blood vessels and leading to microaneurysms. The retinal images give detailed information about the health status of the visual system. Analysis of retinal image is important for an understanding of the stages of Diabetic retinopathy. Microaneurysms observed that appear in retina images, usually, the initial visible sign of DR, if detected early and properly treated can prevent DR complications, including blindness. In this research work, an advanced image modal enhancement comprises of a Contrast Limited Adaptive Histogram Equalization (CLAHE), through morphological image, processing technique with final extraction algorithm is proposed. CLAHE is responsible for the detection, and removal of the retinal optical disk. While the microaneurysm initial indicators are detected by using morphological image processing techniques. The extensive evaluation of the proposed advanced model conducted for microaneurysm detection depicts all stages of DR with an increase in the number of data set related to noise in the image. The microaneurysms noise is associated with stage of retina diseases as well as its early possible diagnosis. Evaluation is also conducted against the proposed model to measure its performance in terms of accuracy, sensitivity as well as specificity in real-time. The results show the test image attained 99.7% accuracy for a real-time database that is better compared with anty colony-based method. A sensitivity of 81% with a specificity of 90% was achieved for the detection of microaneurysms for the e-optha database. The detection of several microaneurysms correlates with stages of DR that prove an analysis of detecting its different stages. As well as it reaches our goal of early detection of DR with high performance in accuracy.

@article{sanya2021novel, abstract = {Diabetic retinopathy (DR) is a common complication of diabetes mellitus and can lead to irreversible blindness. To date, DR is the leading cause of blindness and visual impairment among working adults globally. However, this blindness can be prevented if DR is detected early. Diabetes mellitus slowly affects the retina by damaging retinal blood vessels and leading to microaneurysms. The retinal images give detailed information about the health status of the visual system. Analysis of retinal image is important for an understanding of the stages of Diabetic retinopathy. Microaneurysms observed that appear in retina images, usually, the initial visible sign of DR, if detected early and properly treated can prevent DR complications, including blindness. In this research work, an advanced image modal enhancement comprises of a Contrast Limited Adaptive Histogram Equalization (CLAHE), through morphological image, processing technique with final extraction algorithm is proposed. CLAHE is responsible for the detection, and removal of the retinal optical disk. While the microaneurysm initial indicators are detected by using morphological image processing techniques. The extensive evaluation of the proposed advanced model conducted for microaneurysm detection depicts all stages of DR with an increase in the number of data set related to noise in the image. The microaneurysms noise is associated with stage of retina diseases as well as its early possible diagnosis. Evaluation is also conducted against the proposed model to measure its performance in terms of accuracy, sensitivity as well as specificity in real-time. The results show the test image attained 99.7% accuracy for a real-time database that is better compared with anty colony-based method. A sensitivity of 81% with a specificity of 90% was achieved for the detection of microaneurysms for the e-optha database. The detection of several microaneurysms correlates with stages of DR that prove an analysis of detecting its different stages. As well as it reaches our goal of early detection of DR with high performance in accuracy.}, added-at = {2021-08-26T09:29:55.000+0200}, author = {Sanya, Waheed and Bajpai, Gaurav and Amour, Maryam and Alawi, Mahmoud and Adnan, Ali}, biburl = {https://www.bibsonomy.org/bibtex/272b1ff253906569e95bf282a0f872792/cscjournals}, interhash = {691307f8d947b9b2159845cb7887126e}, intrahash = {72b1ff253906569e95bf282a0f872792}, issn = {1985-2304}, journal = {International Journal of Image Processing (IJIP)}, keywords = {CLAHE, Diabetic Disc, Image, Mathematical Microaneurysms Morphological Operation, Optic Retina Retinopathy,}, language = {English}, month = {August}, number = 3, pages = {22-36}, timestamp = {2021-08-26T09:29:55.000+0200}, title = {A Novel Advanced Approach Using Morphological Image Processing Technique for Early Detection of Diabetes Retinopathy}, url = {https://www.cscjournals.org/library/manuscriptinfo.php?mc=IJIP-1218}, volume = 15, year = 2021 }