1
|
Nagasawa T, Tabuchi H, Masumoto H, Enno H, Niki M, Ohsugi H, Mitamura Y. Accuracy of deep learning, a machine learning technology, using ultra-wide-field fundus ophthalmoscopy for detecting idiopathic macular holes. PeerJ 2018; 6:e5696. [PMID: 30370184 PMCID: PMC6201738 DOI: 10.7717/peerj.5696] [Citation(s) in RCA: 25] [Impact Index Per Article: 4.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/26/2018] [Accepted: 09/05/2018] [Indexed: 11/20/2022] Open
Abstract
We aimed to investigate the detection of idiopathic macular holes (MHs) using ultra-wide-field fundus images (Optos) with deep learning, which is a machine learning technology. The study included 910 Optos color images (715 normal images, 195 MH images). Of these 910 images, 637 were learning images (501 normal images, 136 MH images) and 273 were test images (214 normal images and 59 MH images). We conducted training with a deep convolutional neural network (CNN) using the images and constructed a deep-learning model. The CNN exhibited high sensitivity of 100% (95% confidence interval CI [93.5-100%]) and high specificity of 99.5% (95% CI [97.1-99.9%]). The area under the curve was 0.9993 (95% CI [0.9993-0.9994]). Our findings suggest that MHs could be diagnosed using an approach involving wide angle camera images and deep learning.
Collapse
Affiliation(s)
- Toshihiko Nagasawa
- Department of Ophthalmology, Tsukazaki Hospital, Himeji City, Hyogo Prefecture, Japan
| | - Hitoshi Tabuchi
- Department of Ophthalmology, Tsukazaki Hospital, Himeji City, Hyogo Prefecture, Japan
| | - Hiroki Masumoto
- Department of Ophthalmology, Tsukazaki Hospital, Himeji City, Hyogo Prefecture, Japan
| | | | - Masanori Niki
- Department of Ophthalmology, Institute of Biomedical Sciences, Tokushima University, Tokushima City, Tokushima Prefecture, Japan
| | - Hideharu Ohsugi
- Department of Ophthalmology, Tsukazaki Hospital, Himeji City, Hyogo Prefecture, Japan
| | - Yoshinori Mitamura
- Department of Ophthalmology, Institute of Biomedical Sciences, Tokushima University, Tokushima City, Tokushima Prefecture, Japan
| |
Collapse
|
2
|
Zhu W, Zhang L, Shi F, Xiang D, Wang L, Guo J, Yang X, Chen H, Chen X. Automated framework for intraretinal cystoid macular edema segmentation in three-dimensional optical coherence tomography images with macular hole. JOURNAL OF BIOMEDICAL OPTICS 2017; 22:76014. [PMID: 28732095 DOI: 10.1117/1.jbo.22.7.076014] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/11/2017] [Accepted: 07/05/2017] [Indexed: 06/07/2023]
Abstract
Cystoid macular edema (CME) and macular hole (MH) are the leading causes for visual loss in retinal diseases. The volume of the CMEs can be an accurate predictor for visual prognosis. This paper presents an automatic method to segment the CMEs from the abnormal retina with coexistence of MH in three-dimensional-optical coherence tomography images. The proposed framework consists of preprocessing and CMEs segmentation. The preprocessing part includes denoising, intraretinal layers segmentation and flattening, and MH and vessel silhouettes exclusion. In the CMEs segmentation, a three-step strategy is applied. First, an AdaBoost classifier trained with 57 features is employed to generate the initialization results. Second, an automated shape-constrained graph cut algorithm is applied to obtain the refined results. Finally, cyst area information is used to remove false positives (FPs). The method was evaluated on 19 eyes with coexistence of CMEs and MH from 18 subjects. The true positive volume fraction, FP volume fraction, dice similarity coefficient, and accuracy rate for CMEs segmentation were 81.0%±7.8%, 0.80%±0.63%, 80.9%±5.7%, and 99.7%±0.1%, respectively.
Collapse
Affiliation(s)
- Weifang Zhu
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Li Zhang
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Fei Shi
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Dehui Xiang
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Lirong Wang
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Jingyun Guo
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Xiaoling Yang
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| | - Haoyu Chen
- Shantou University and the Chinese University of Hong Kong, Joint Shantou International Eye Center, Shantou, ChinacThe Chinese University of Hong Kong, Department of Ophthalmology and Visual Sciences, Hong Kong, China
| | - Xinjian Chen
- Soochow University, School of Electronics and Information Engineering, Suzhou, China
| |
Collapse
|
3
|
Dihowm F, MacCumber M. Comparison of outcomes between 20, 23 and 25 gauge vitrectomy for idiopathic macular hole. Int J Retina Vitreous 2015; 1:6. [PMID: 27847599 PMCID: PMC5066517 DOI: 10.1186/s40942-015-0007-6] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/22/2015] [Accepted: 05/26/2015] [Indexed: 11/10/2022] Open
Abstract
Purpose To compare the results of 20, 23, 25 gauge pars plana vitrectomy (PPV) with two different gas tamponades for idiopathic macular hole (MH) in a multi-surgeon vitroretinal practice. Methods In this comparative, retrospective, interventional case series, the medical charts of 142 eyes/130 patients were reviewed. Patients who matched our inclusion criteria: eye with stage 2, 3, or 4 MH that underwent 20, 23, or 25 gauge PPV, internal limiting membrane (ILM) peeling, and fluid-gas exchange from January, 2005 to May, 2012 and had at least 6 months follow-up. The best current corrected visual acuity (VA) and anatomical status of the MH were assessed by optical coherent tomography (OCT) at 6 months, 1 year, and 2 years after vitrectomy. Results The MH closed successfully after primary vitrectomy in 86.5 % (20 gauge), 96.4 % (23 gauge), and 92 % (25 gauge). Preoperative VA median were 20\126 (20 gauge), 20\100 (23 gauge), and 20\80 (25 gauge). At 6 months and 2 years postoperative VA did not differ significantly between the 3 groups (p = 0.570, and 0.054 respectively). However, at 12 months postoperative VA median 20\60 (20 gauge), 20\69 (23 gauge), and 20\40 (25 gauge) differ significantly (p = 0.005) likely due to cataract changes. The final median postoperative VA (at 2 years) in 25 gauge PPV group was 20/40 which was better than final visual outcomes for 20, and 23 gauge PPV groups (20/50, and 20/55 respectively). The different was not a statistically significant. MH closed successfully in 96 % (C3F8), and 88.1 % (SF6) (p = 0.063). Preoperative median VA was 20/100 in both groups of gas. At 6 months, 1 year, and 2 years postoperative median VAs did not differ significant between the 2 groups (p = 0.076, 0.343, and 0.309 respectively). MH closed successfully in (96.9 %) 12-14 % C3F8, and (95.3 %) 15-16 % C3F8 (p = 0.611). MH closed in (82.1 %) 18-20 % SF6, and (96.4 %) 22-26 % SF6 (p=0.053). Conclusion Based on the results of this study, 20, 23, and 25 gauge of PPV have similar MH closure rates and VA outcomes. SF6 at 22-26 % or C3F8 at 12-14 % achieved maximum closure rates.
Collapse
Affiliation(s)
- Fatma Dihowm
- grid.240684.c0000000107053621Graduate College, Rush University Medical Center, Chicago, IL USA
| | - Mathew MacCumber
- grid.240684.c0000000107053621Department of Ophthalmology, Rush University Medical Center, Chicago, IL USA
| |
Collapse
|
4
|
|
5
|
Liu YY, Ishikawa H, Chen M, Wollstein G, Duker JS, Fujimoto JG, Schuman JS, Rehg JM. Computerized macular pathology diagnosis in spectral domain optical coherence tomography scans based on multiscale texture and shape features. Invest Ophthalmol Vis Sci 2011; 52:8316-22. [PMID: 21911579 DOI: 10.1167/iovs.10-7012] [Citation(s) in RCA: 32] [Impact Index Per Article: 2.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/24/2022] Open
Abstract
PURPOSE To develop an automated method to identify the normal macula and three macular pathologies (macular hole [MH], macular edema [ME], and age-related macular degeneration [AMD]) from the fovea-centered cross sections in three-dimensional (3D) spectral-domain optical coherence tomography (SD-OCT) images. METHODS A sample of SD-OCT macular scans (macular cube 200 × 200 or 512 × 128 scan protocol; Cirrus HD-OCT; Carl Zeiss Meditec, Inc., Dublin, CA) was obtained from healthy subjects and subjects with MH, ME, and/or AMD (dataset for development: 326 scans from 136 subjects [193 eyes], and dataset for testing: 131 scans from 37 subjects [58 eyes]). A fovea-centered cross-sectional slice for each of the SD-OCT images was encoded using spatially distributed multiscale texture and shape features. Three ophthalmologists labeled each fovea-centered slice independently, and the majority opinion for each pathology was used as the ground truth. Machine learning algorithms were used to identify the discriminative features automatically. Two-class support vector machine classifiers were trained to identify the presence of normal macula and each of the three pathologies separately. The area under the receiver operating characteristic curve (AUC) was calculated to assess the performance. RESULTS The cross-validation AUC result on the development dataset was 0.976, 0.931, 0939, and 0.938, and the AUC result on the holdout testing set was 0.978, 0.969, 0.941, and 0.975, for identifying normal macula, MH, ME, and AMD, respectively. CONCLUSIONS The proposed automated data-driven method successfully identified various macular pathologies (all AUC > 0.94). This method may effectively identify the discriminative features without relying on a potentially error-prone segmentation module.
Collapse
Affiliation(s)
- Yu-Ying Liu
- School of Interactive Computing, Georgia Institute of Technology, Atlanta, Georgia, USA
| | | | | | | | | | | | | | | |
Collapse
|
6
|
Liu YY, Chen M, Ishikawa H, Wollstein G, Schuman JS, Rehg JM. Automated macular pathology diagnosis in retinal OCT images using multi-scale spatial pyramid and local binary patterns in texture and shape encoding. Med Image Anal 2011; 15:748-59. [PMID: 21737338 DOI: 10.1016/j.media.2011.06.005] [Citation(s) in RCA: 70] [Impact Index Per Article: 5.4] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/02/2011] [Revised: 06/03/2011] [Accepted: 06/06/2011] [Indexed: 11/18/2022]
Abstract
We address a novel problem domain in the analysis of optical coherence tomography (OCT) images: the diagnosis of multiple macular pathologies in retinal OCT images. The goal is to identify the presence of normal macula and each of three types of macular pathologies, namely, macular edema, macular hole, and age-related macular degeneration, in the OCT slice centered at the fovea. We use a machine learning approach based on global image descriptors formed from a multi-scale spatial pyramid. Our local features are dimension-reduced local binary pattern histograms, which are capable of encoding texture and shape information in retinal OCT images and their edge maps, respectively. Our representation operates at multiple spatial scales and granularities, leading to robust performance. We use 2-class support vector machine classifiers to identify the presence of normal macula and each of the three pathologies. To further discriminate sub-types within a pathology, we also build a classifier to differentiate full-thickness holes from pseudo-holes within the macular hole category. We conduct extensive experiments on a large dataset of 326 OCT scans from 136 subjects. The results show that the proposed method is very effective (all AUC>0.93).
Collapse
Affiliation(s)
- Yu-Ying Liu
- School of Interactive Computing, College of Computing, Georgia Institute of Technology, Atlanta, GA, USA.
| | | | | | | | | | | |
Collapse
|
7
|
Abstract
PURPOSE To assess patient expectation, visual improvement, and patient satisfaction after macular hole surgery. METHODS Fifty-three consecutive patients completed a self-administered questionnaire after macular hole surgery. Responses for expectation of visual improvement before surgery and subjective visual assessment after surgery were scored from 0 to 10 (0 = no improvement in vision, 10 = full improvement). Visual acuity was measured using Snellen charts and converted to logMAR equivalent for statistical analysis. Wilcoxon and Mann-Whitney U tests were performed using the Minitab statistical software program; p < 0.05 was considered significant. RESULTS Thirty-eight (71%) patients were satisfied with the outcome, 14 (26%) were not, and 1 was uncertain. Both satisfied and dissatisfied patients had no difference in baseline preoperative best-corrected visual acuity (BCVA; 6/60 and 6/60, respectively). There was also no significant difference in median postoperative BCVA between both groups (6/24 and 6/29, respectively, p = 0.6). In satisfied patients, median vision expectation score of 6 was equally similar to the median subjective postoperative vision improvement score of 6, p = 0.7. However, in dissatisfied patients, median vision expectation score (8) was significantly higher than median postoperative subjective vision improvement score (1), p = 0.0001. The commonest reason for dissatisfaction was perceived visual reduction (7/14 patients). CONCLUSIONS In the dissatisfied patients, expectation was significantly higher, and subjective assessment of vision after surgery was lower than in satisfied patients, although they had similar improvement in BCVA. Visual acuity and patient satisfaction are not always directly correlated. An important factor in patient satisfaction is expectation.
Collapse
|
8
|
Sen P, Bhargava A, Vijaya L, George R. Prevalence of idiopathic macular hole in adult rural and urban south Indian population. Clin Exp Ophthalmol 2008; 36:257-60. [PMID: 18412595 DOI: 10.1111/j.1442-9071.2008.01715.x] [Citation(s) in RCA: 40] [Impact Index Per Article: 2.5] [Reference Citation Analysis] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/03/2023]
Affiliation(s)
- Parveen Sen
- Bhagwan Mahavir Vitreo-Retinal Services, Medical Research Foundation, Sankara Nethralaya, Chennai, India
| | | | | | | |
Collapse
|
9
|
Mönestam E, Andersson M, Aström S. Associations between subjective and objective visual function in patients with unilateral macular holes. Vision Res 2007; 48:104-8. [PMID: 18061643 DOI: 10.1016/j.visres.2007.10.013] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/25/2007] [Revised: 10/17/2007] [Accepted: 10/25/2007] [Indexed: 10/22/2022]
Abstract
Forty-six patients with uniocular macular holes and unaffected, fellow eyes were studied to evaluate inter- and intraocular associations between various objective tests of visual function and perceived visual ability. The affected eye had significant associations between visual acuity (VA) and the fovea threshold test, but for the fellow eye only VA and low-contrast VA 10% were associated. The reduction in visual acuity under low-contrast conditions relative to high-contrast did not differ between the affected eye and the healthy eye. Subjective visual ability seems to depend more on the visual acuity of the affected eye than the healthy eye.
Collapse
Affiliation(s)
- E Mönestam
- Department of Clinical Sciences/Ophthalmology, Umeå University, S-90185 Umeå, Sweden.
| | | | | |
Collapse
|
10
|
Mireskandari K, Garnham L, Sheard R, Ezra E, Gregor ZJ, Sloper JJ. A prospective study of the effect of a unilateral macular hole on sensory and motor binocular function and recovery following successful surgery. Br J Ophthalmol 2004; 88:1320-4. [PMID: 15377559 PMCID: PMC1772365 DOI: 10.1136/bjo.2004.042093] [Citation(s) in RCA: 22] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/04/2022]
Abstract
AIMS To examine the effect of a unilateral full thickness macular hole on sensory and motor binocular function and to study recovery after successful surgical closure. METHODS Twenty eight consecutive patients undergoing surgery for a unilateral macular hole underwent orthoptic examination, including measurements of Titmus and TNO stereoacuity and motor fusion range before surgery. Twenty three patients had successful anatomical closure. Fifteen of these patients, who had both improved acuity in the operated eye following surgery and were available for further testing, underwent repeat orthoptic assessment 2-7 months after surgery. RESULTS In all patients stereoacuity was reduced before surgery, but few patients were subjectively aware of a deficit of depth perception affecting their everyday life. In those patients with improved Snellen acuity after surgery, stereoacuity measured by the Titmus stereotest also improved significantly, but not that measured by the TNO test. Two patients were aware of a subjective improvement in depth perception. Motor fusion was markedly reduced compared to normal before surgery, with only limited recovery after surgery. CONCLUSION A unilateral macular hole notably reduced both stereoacuity and motor fusion. Successful closure improved the deficit in stereoacuity associated with the hole when measured by a stereotest using contoured stimuli. The majority of patients were not subjectively aware of the deficit in stereoacuity or its improvement following surgery.
Collapse
Affiliation(s)
- K Mireskandari
- Vitreo-retinal service, Moorfields Eye Hospital, City Road, London EC1V 2PD, UK
| | | | | | | | | | | |
Collapse
|
11
|
Casuso LA, Scott IU, Flynn HW, Gass JD, Smiddy WE, Lewis ML, Schiffman J. Long-term follow-up of unoperated macular holes. Ophthalmology 2001; 108:1150-5. [PMID: 11382645 DOI: 10.1016/s0161-6420(01)00581-4] [Citation(s) in RCA: 64] [Impact Index Per Article: 2.8] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/16/2022] Open
Abstract
OBJECTIVE To investigate the anatomic and visual acuity outcomes among patients with unoperated macular holes and at least 5 years of follow-up. DESIGN Retrospective, noncomparative case series from an institutional practice setting. PARTICIPANTS All patients with unoperated full-thickness macular holes evaluated at Bascom Palmer Eye Institute between January 1, 1968 and December 31, 1993 and observed for at least 5 years. METHODS Demographic and clinical data were abstracted from patients' medical records and ophthalmologic photography records. For patients with bilateral macular holes, only one eye was included. MAIN OUTCOME MEASURES Visual acuity and clinical features on initial examination, at 5 years, and at final follow-up. RESULTS The study included 65 eyes of 65 patients with a median age of 65 years (range, 52-85 years) and a median follow-up of 9.3 years (range, 5-29 years). On initial examination at Bascom Palmer Eye Institute, the macular hole was stage 2 in 15 eyes (24%), stage 3 in 23 eyes (37%), and stage 4 in 25 eyes (40%). At final follow-up, the macular hole was stage 3 in 10 eyes (16%) and stage 4 in 53 eyes (84%). Visual acuity was 20/200 or worse in 35 eyes (54%) on initial examination, in 43 eyes (74%) at 5 years, and in 53 eyes (82%) at final follow-up. Poorer visual acuity on initial examination was a significant predictor of poorer final vision (P < 0.01). Other accompanying clinical features such as the presence of operculum, posterior vitreous detachment, and epiretinal membrane were not significantly associated with final vision. Throughout follow-up, there was a redistribution and reduced number of yellow nodular opacities at the level of the retinal pigment epithelium at the base of the macular holes and the development of retinal pigment epithelial atrophy around the macular holes. CONCLUSIONS Long-term follow-up of unoperated macular holes demonstrates progression in hole size and stage, vision loss which generally stabilizes at the 20/200 to 20/400 level, a redistribution and reduced number of yellow nodular opacities at the level of the retinal pigment epithelium, and the development of retinal pigment epithelial atrophy surrounding the macular hole, resulting in a "bull's-eye" macular appearance.
Collapse
Affiliation(s)
- L A Casuso
- Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida 33136, USA
| | | | | | | | | | | | | |
Collapse
|