scholarly journals Glaucoma Diagnosis with Machine Learning Based on Optical Coherence Tomography and Color Fundus Images

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Guangzhou An ◽  
Kazuko Omodaka ◽  
Kazuki Hashimoto ◽  
Satoru Tsuda ◽  
Yukihiro Shiga ◽  
...  
Keyword(s):  
Machine Learning ◽  
Optical Coherence Tomography ◽  
Optic Disc ◽  
Learning System ◽  
Ganglion Cell Complex ◽  
Fundus Images ◽  
Optical Coherence ◽  
Glaucoma Diagnosis ◽  
Rnfl Thickness ◽  
Color Fundus Images

This study aimed to develop a machine learning-based algorithm for glaucoma diagnosis in patients with open-angle glaucoma, based on three-dimensional optical coherence tomography (OCT) data and color fundus images. In this study, 208 glaucomatous and 149 healthy eyes were enrolled, and color fundus images and volumetric OCT data from the optic disc and macular area of these eyes were captured with a spectral-domain OCT (3D OCT-2000, Topcon). Thickness and deviation maps were created with a segmentation algorithm. Transfer learning of convolutional neural network (CNN) was used with the following types of input images: (1) fundus image of optic disc in grayscale format, (2) disc retinal nerve fiber layer (RNFL) thickness map, (3) macular ganglion cell complex (GCC) thickness map, (4) disc RNFL deviation map, and (5) macular GCC deviation map. Data augmentation and dropout were performed to train the CNN. For combining the results from each CNN model, a random forest (RF) was trained to classify the disc fundus images of healthy and glaucomatous eyes using feature vector representation of each input image, removing the second fully connected layer. The area under receiver operating characteristic curve (AUC) of a 10-fold cross validation (CV) was used to evaluate the models. The 10-fold CV AUCs of the CNNs were 0.940 for color fundus images, 0.942 for RNFL thickness maps, 0.944 for macular GCC thickness maps, 0.949 for disc RNFL deviation maps, and 0.952 for macular GCC deviation maps. The RF combining the five separate CNN models improved the 10-fold CV AUC to 0.963. Therefore, the machine learning system described here can accurately differentiate between healthy and glaucomatous subjects based on their extracted images from OCT data and color fundus images. This system should help to improve the diagnostic accuracy in glaucoma.

scholarly journals Effect of trabeculectomy on RNFL thickness and optic disc parameters using optical coherence tomography

Eye ◽  
2012 ◽  
Vol 26 (8) ◽  
pp. 1131-1137 ◽  
Author(s):  
N Raghu ◽  
S S Pandav ◽  
S Kaushik ◽  
P Ichhpujani ◽  
A Gupta
Keyword(s):  
Optical Coherence Tomography ◽  
Optic Disc ◽  
Optical Coherence ◽  
Rnfl Thickness

scholarly journals Automated Segmentation of Optic Disc and Cup in Color Fundus Images

2019 ◽  
Vol 2 (1) ◽  
Author(s):  
Vijay M Mane
Keyword(s):  
Optic Disc ◽  
Disease Diagnosis ◽  
Eye Disease ◽  
Automated Segmentation ◽  
Fundus Images ◽  
Average Sensitivity ◽  
Watershed Transformation ◽  
Glaucoma Diagnosis ◽  
Optic Cup ◽  
Color Fundus Images

An automatic Optic disc and Optic cup detection technique which is an important step in developing systems for computer-aided eye disease diagnosis is presented in this paper. This paper presents an algorithm for localization and segmentation of optic disc from digital retinal images. OD localization is achieved by circular Hough transform using morphological preprocessing and segmentation is achieved by watershed transformation. Optic cup segmentation is achieved by marker controlled watershed transformation. The optic disc to cup ratio (CDR) is calculated which is an important parameter for glaucoma diagnosis. The presented algorithm is evaluated against publically available DRIVE dataset. The presented methodology achieved 88% average sensitivity and 80% average overlap. The average CDR detected is 0.1983.

scholarly journals Influence of Disc Size on the Diagnostic Accuracy of Cirrus Spectral-Domain Optical Coherence Tomography in Glaucoma

2018 ◽  
Vol 2018 ◽  
pp. 1-5 ◽  
Author(s):  
Shino Sato ◽  
Kaori Ukegawa ◽  
Eri Nitta ◽  
Kazuyuki Hirooka
Keyword(s):  
Optical Coherence Tomography ◽  
Diagnostic Accuracy ◽  
Optic Disc ◽  
Characteristic Curve ◽  
Spectral Domain ◽  
Optical Coherence ◽  
Chi Square ◽  
Optic Disc Size ◽  
Glaucoma Diagnosis ◽  
Disc Size

Purpose. To examine the influence of optic disc size on the diagnostic accuracy of optic nerve head (ONH) parameters determined by Cirrus spectral-domain optical coherence tomography (Cirrus HD-OCT). Methods. A total of 51 eyes of 51 normal participants and 71 eyes of 71 glaucoma patients were examined. ONH imaging was obtained by Cirrus HD-OCT. Sensitivity at a fixed 90% specificity along with the area under the receiver operating characteristic curve (AUC) for continuous parameters were analyzed. We also examined the coefficients of variation (CoV) for sensitivity estimates, as these have been used to test and quantify the influence of optic disc size on diagnostic accuracy. The influence of optic disc size on the glaucoma diagnosis was assessed by the likelihood ratio chi-square test. Results. Among the continuous parameters, the best diagnostic accuracy was seen for the average rim area, which had an AUC of 0.96. The most reliable factor across the disc size groups was the rim area (CoV, 2.8%). The diagnostic accuracy of the rim area did not appear to be influenced by optic disc size (P=0.17). Conclusions. The high diagnostic accuracy of the rim area demonstrated by Cirrus HD-OCT for the quantitative assessment of the ONH was not significantly affected by disc size in this study.

scholarly journals Thinner temporal peripapillary retinal nerve fibre layer in Stargardt disease detected by optical coherence tomography

2020 ◽  
Author(s):  
Michael Reich ◽  
Jan Lübke ◽  
Lutz Joachimsen ◽  
Julia Stifter ◽  
Sebastian Küchlin ◽  
...  
Keyword(s):  
Optical Coherence Tomography ◽  
Optic Disc ◽  
Nerve Fibre ◽  
Retinal Nerve Fibre Layer ◽  
Optical Coherence ◽  
Stargardt Disease ◽  
Rnfl Thickness ◽  
Fibre Layer ◽  
Manual Adjustment ◽  
Nerve Fibre Layer

Abstract Purpose To evaluate peripapillary retinal nerve fibre layer (RNFL) thickness measured by spectral domain optical coherence tomography (OCT) in patients with Stargardt disease (STGD). Methods A cross-sectional, monocentric, observational case-control study. Twenty patients (39 eyes) with ABCA4 mutations graded according to the Fishman STGD classification were included. RNFL measurement was performed using Heidelberg Spectralis SD-OCT. RNFL thickness in STGD patients was compared to age-matched data of healthy individuals provided by the device’s manufacturer. A manual readjustment of the optic disc-fovea angle was performed when needed. Results The mean age at first diagnosis of STGD was 22.9 years (range 9 to 50) and 39.1 years (range 18 to 74) at the time of examination. Thirty-nine percent of eyes (15 eyes) needed manual adjustment of the optic disc-fovea angle due to malfixation of the patients during OCT. The temporal quadrant corresponding to the macula showed a RNFL 16% thinner than controls (mean − 12 μm, 95%CI − 9 to −15 μm). However, global RNFL thickness did not differ from controls due to increased RNFL thickness of 12% in the nasal sectors. Duration and stage of STGD were not correlated to thinner RNFL. Conclusion STGD seems to be associated with thinner peripapillary RNFL in the sector of axons projecting to the degenerated macular area. It is yet unclear as to whether this results from anterograde transneuronal degeneration of direct injury to retinal ganglion cells.

scholarly journals Ganglion cell complex analysis in thalassemia major patients measured by optical coherence tomography

2021 ◽  
Vol 1 (2) ◽  
pp. 63-68
Author(s):  
Semra Koca ◽  
Selin Yakarisik
Keyword(s):  
Optical Coherence Tomography ◽  
Ganglion Cell ◽  
Complex Analysis ◽  
Ferritin Level ◽  
Thalassemia Major ◽  
Cell Complex ◽  
Diagnostic Methods ◽  
Ganglion Cell Complex ◽  
Optical Coherence ◽  
Rnfl Thickness

To analyze the changes in ganglion cell complex (GCC), peripapillary retinal nerve fiber layer (RNFL) thickness and central macular thickness (CMT) on spectral domain optical coherence tomography (OCT) in patients with thalassemia major. Forty one eyes of 41 patients with thalassemia major and 41 eyes of 41 healthy subjects were included in this prospective and comparative study. Peripapillary RNFL thickness, CMT and macular GCC thickness were evaluated with OCT (Cirrus HD-OCT 5000Carl Zeiss Meditec, Inc, Dublin, CA, USA) in all patients and healthy controls. Additionally, disease duration, serum ferritin level, hemoglobin concentration, the dosage and duration of chelation therapy, count of transfusion, patient’s weight were analyzed in thalassemia major group. RNFL thickness values were lower in the thalassemia patients but the difference was not statistically significant (except superior quadrant) and there was no significant differences in the mean CMT measurements. GCC thickness was thinner in all areas ( average, superior, inferior, superior-temporal, inferior-temporal, superior-nasal, inferior-nasal) but only the thinning in the inferior-temporal was statistically significant. GCC and RNFL thickness changes occur earlier than CMT changes in β-thalassemia major patients. GCC thickness measurements can be used for follow-up in combination with other diagnostic methods.

scholarly journals Optic Disc and Macula Localization from Retinal Optical Coherence Tomography and Fundus Image

2018 ◽  
Vol 8 (6) ◽  
pp. 5050
Author(s):  
Rodiah Rodiah ◽  
Sarifuddin Madenda ◽  
Diana Tri Susetianigtias ◽  
Dewi Agushinta Rahayu ◽  
Ety Sutanty
Keyword(s):  
Optical Coherence Tomography ◽  
Optic Disc ◽  
Fundus Images ◽  
Optical Coherence ◽  
Fundus Image ◽  
Optical Disc

<p>This research used images from Optical Coherence Tomography (OCT) examination as well as fundus images to localize the optical disc and macular layer of retina. The researchers utilized the OCT and fundus image to interpret the distance between macular center and optic disc in the image. The distance will express the area of macula that can be employed for further research. This distance could recognize the thickness of macula parameters diameter that will be used in localizing process of optic disc and macula. The parameters are the circle radius, the size of window’s filter, the constant value and the size of optic disc element structure as well as the size of macula. The results of this study are expected to improve the accuracy of macula detection that experience the edema.</p>

Accuracy of optical coherence tomography for diagnosing glaucoma: an overview of systematic reviews

2020 ◽  
pp. bjophthalmol-2020-316152 ◽  
Author(s):  
Manuele Michelessi ◽  
Tianjing Li ◽  
Alba Miele ◽  
Augusto Azuara-Blanco ◽  
Riaz Qureshi ◽  
...  
Keyword(s):  
Optical Coherence Tomography ◽  
Systematic Reviews ◽  
High Specificity ◽  
Retinal Nerve Fibre Layer ◽  
Risk Of Bias ◽  
Optical Coherence ◽  
Glaucoma Diagnosis ◽  
Rnfl Thickness ◽  
Overview Of Systematic Reviews

AimsTo assess the diagnostic accuracy (DTA) of optical coherence tomography (OCT) for detecting glaucoma by systematically searching and appraising systematic reviews (SRs) on this issue.MethodsWe searched a database of SRs in eyes and vision maintained by the Cochrane Eyes and Vision United States on the DTA of OCT for detecting glaucoma. Two authors working independently screened the records, abstracted data and assessed the risk of bias using the Risk of Bias in Systematic Reviews checklist. We extracted quantitative DTA estimates as well as qualitative statements on their relevance to practice.ResultsWe included four SRs published between 2015 and 2018. These SRs included between 17 and 113 studies on OCT for glaucoma diagnosis. Two reviews were at low risk of bias and the other two had two to four domains at high or unclear risk of bias with concerns on applicability. The two reliable SRs reported the accuracy of average retinal nerve fibre layer (RNFL) thickness and found a sensitivity of 0.69 (0.63 to 0.73) and 0.78 (0.74 to 0.83) and a specificity of 0.94 (0.93 to 0.95) and 0.93 (0.92 to 0.95) in 57 and 50 studies, respectively. Only one review included a clear specification of the clinical pathway. Both reviews highlighted the limitations of primary DTA studies on this topic.ConclusionsThe quality of published DTA reviews on OCT for diagnosing glaucoma was mixed. Two reliable SRs found moderate sensitivity at high specificity for average RNFL thickness in diagnosing manifest glaucoma. Our overview suggests that the methodological quality of both primary and secondary DTA research on glaucoma is in need of improvement.

scholarly journals Development and Validation of a Deep Learning System for Diagnosing Glaucoma Using Optical Coherence Tomography

2020 ◽  
Vol 9 (7) ◽  
pp. 2167
Author(s):  
Ko Eun Kim ◽  
Joon Mo Kim ◽  
Ji Eun Song ◽  
Changwon Kee ◽  
Jong Chul Han ◽  
...  
Keyword(s):  
Optical Coherence Tomography ◽  
Deep Learning ◽  
External Validation ◽  
Learning System ◽  
Validation Dataset ◽  
Inner Plexiform Layer ◽  
Optical Coherence ◽  
Diagnostic Ability ◽  
Rnfl Thickness ◽  
Validation Set

This study aimed to develop and validate a deep learning system for diagnosing glaucoma using optical coherence tomography (OCT). A training set of 1822 eyes (332 control, 1490 glaucoma) with 7288 OCT images, an internal validation set of 425 eyes (104 control, 321 glaucoma) with 1700 images, and an external validation set of 355 eyes (108 control, 247 glaucoma) with 1420 images were included. Deviation and thickness maps of retinal nerve fiber layer (RNFL) and ganglion cell–inner plexiform layer (GCIPL) analyses were used to develop the deep learning system for glaucoma diagnosis based on the visual geometry group deep convolutional neural network (VGG-19) model. The diagnostic abilities of deep learning models using different OCT maps were evaluated, and the best model was compared with the diagnostic results produced by two glaucoma specialists. The glaucoma-diagnostic ability was highest when the deep learning system used the RNFL thickness map alone (area under the receiver operating characteristic curve (AUROC) 0.987), followed by the RNFL deviation map (AUROC 0.974), the GCIPL thickness map (AUROC 0.966), and the GCIPL deviation map (AUROC 0.903). Among combination sets, use of the RNFL and GCIPL deviation map showed the highest diagnostic ability, showing similar results when tested via an external validation dataset. The inclusion of the axial length did not significantly affect the diagnostic performance of the deep learning system. The location of glaucomatous damage showed generally high level of agreement between the heatmap and the diagnosis of glaucoma specialists, with 90.0% agreement when using the RNFL thickness map and 88.0% when using the GCIPL thickness map. In conclusion, our deep learning system showed high glaucoma-diagnostic abilities using OCT thickness and deviation maps. It also showed detection patterns similar to those of glaucoma specialists, showing promising results for future clinical application as an interpretable computer-aided diagnosis.

Sign in / Sign up

Export Citation Format

Share Document