Portal de Periódicos da CAPES

A Review on Glaucoma Disease Detection Using Computerized Techniques

2021; Institute of Electrical and Electronics Engineers; Volume: 9; Linguagem: Inglês

10.1109/access.2021.3061451

2169-3536

Faizan Abdullah, Rakhshanda Imtiaz, Hussain Ahmad Madni, Haroon Ahmed Khan, Tariq M. Khan, Mohammad A. U. Khan, Syed S. Naqvi,

Glaucoma and retinal disorders

Glaucoma is an incurable eye disease that leads to slow progressive degeneration of the retina.It cannot be fully cured, however, its progression can be controlled in case of early diagnosis.Unfortunately, due to the absence of clear symptoms during the early stages, early diagnosis are rare.Glaucoma must be detected at early stages since late diagnosis can lead to permanent vision loss.Glaucoma affects the retina by damaging the Optic Nerve Head (ONH).Its diagnosis is dependent on the measurements of Optic Cup (OC) and Optic Disc (OD) in the retina.Computer vision techniques have been shown to diagnose glaucoma effectively and correctly with little overhead.These techniques measure OC and OC dimensions using machine learning based classification and segmentation algorithms.This article aims to provide a comprehensive overview of various existing techniques that use machine learning to detect and diagnose glaucoma based on fundus images.Readers would be able to understand the challenges glaucoma presents from an image processing and machine learning stand-point and will be able to identify gaps in current research.INDEX TERMS Glaucoma, convolutional neural networks (CNN), diabetic retinopathy, cup-to-disc ratio (CDR), optic nerve head (ONH), optic cup (OC), optic disc (OD), intra ocular pressure (IOP). II. DATASETSDifferent datasets used in the Glaucoma detection techniques have been described and grouped as follows. A. SINDIA study was conducted on Indian population for the analysis of both eyes.People with age range from 40 to 40 to 83 years, were included in the analysis.Previous surgery character was taken into account for the analysis.This dataset contains 5670 normal and 113 Glaucomatous images. B. SCESThis cross-sectional study took place in Singapore on the basis of population in which 1060 chinese took part.All subjects went through the Optical Coherence Tomography (OCT).This dataset is comprised of 1630 normal images and 46 Glaucomatous images with a sum of 1676 images.This dataset is subjected to classify the images as normal or Glaucomatous. C. SIMESFor the SiMES, a study was conducted to analyse eye diseases especially in adults living in Malays Singapore.People with age range of 40 to 79 were taken into consideration for the analysis.Patients were assessed by retinal photography, optic disc, ocular biometry and digital lens.The dataset consists of 482 as normal images and 168 as Glaucomatous iamges used for the purpose of classification. D. ARIAThis dataset is mainly focused on efficient measurement and detection of retinal vessels which can be implemented on both high and low resolution fundus images.This dataset contains 161 images with resolution of 768×576 used for the detection and measurement of retinal vessels and analysis of Glaucoma in the eye. E. DRISHTI-GSThe Drishti-GS dataset consists of 101 retinal images which are attained from the Aravind eye hospital, India. VOLUME 9, 2021The resolution of these retinal image is 2896 × 1944 where Field of View (FOV) is 30 degree.The age of the patients were in the range of 40 and 80.The groundtruths of OD and OC exist in the dataset.Moreover, the images were annotated by 4 ophthalmologists. F. RIM-ONEIn RIM-ONE v3, 159 retinal fundus images are present along with their groundtruths that have been annotated by the ophthalmologists.This dataset is comprised of 74 glaucomatous and 85 non glaucomatous images. G. RIGARIGA is a dataset used for the diagnosis of Glaucoma and it stands for Retinal Images for Glaucoma Analysis.This dataset consists of 750 retinal fundus images.These retinal fundus images are acquired from three different resources which are MESSIDOR, Magrabi Eye Centre in Riyadh and Bin Rushed ophthalmic centre in Riyadh.This dataset contains glaucomatous as wells as non-glacomatous images along with their groundtruths that are manually annotated by six ophthalmologists. H. ORIGA-LIGHTThe retinal fundus images present in the ORIGA-light dataset are collected by the Singapore Malay Eye Study (SiMES) [6].The process of collection was funded by National Medical Research Council and conducted by the Singapore Eye Research Institute which was completed in the duration of 3 years.This dataset provides the assistance to the researcher for the segmentation of retinal images that helps in the analysis of Glaucoma.This dataset also contains the groundtruths that facilitate the researcher and provides the benchmark for the evaluation of the tools that are designed for the diagnosis of Glaucoma.To study this case, retinal fundus images of both eyes were taken, and the age of the people that were examined, was between 40 and 80.The number images that were kept for making this dataset are 650 in total.Out of 650 images, there are 168 images that are Glaucomatous and other 482 images are nonglaucomatous.This dataset is comprised of retinal fundus images along with their groundtruths.The trained professionals of Singapore Eye Research Institute have segmented and annotated these 650 images that exist in the ORIGA-light dataset. I. ACRIMAThis dataset consists of 705 retinal images with 396 glaucomatous and 309 healthy images.The retinal images were taken from both eye (i.e.left and right), and were formerly dilated and centred in the OD.The Topcon TRC retinal camera and IMAGEnet capture system were used to capture the retinal images.The FOV of retinal images is 35 degree while resolution is 2048 × 1536 pixels.The images in this dataset were annotated by two specialists of glaucoma at the Fundación Oftalmológica del Mediterráneo (FOM).To label these images, no other clinical information was used.In the first version of this dataset, annotations of OD and OC are not given, therefore, it can only be employed for classification purpose.