Segmentation technique and dynamic ensemble selection to enhance glaucoma severity detection

The severity of glaucoma can be observed by categorising glaucoma diseases into several classes based on a classification process. The two most suitable parameters, cup-to-disc ratio (CDR) and peripapillary atrophy (PPA), which are commonly used to identify glaucoma are utilized in this study to strengthen the classification. First, an active contour snake (ACS) is employed to retrieve both optic disc (OD) and optic cup (OC) values, which are required to calculate the CDR. Moreover, Otsu segmentation and thresholding techniques are used to identify PPA, and the features are then extracted using a grey-level co-occurrence matrix (GLCM). An advanced segmentation technique, combined with an improved classifier called dynamic ensemble selection (DES), is proposed to classify glaucoma. Because DES is generally used to handle an imbalanced dataset, the proposed model is expected to detect glaucoma severity and determine the subsequent treatment accurately. The proposed model obtains a higher mean accuracy (0.96) than the deep learning-based U-Net (0.90) when evaluated using three datasets of 250 retinal fundus images (200 training, 50 testings) based on the 5-fold cross-validation scheme.

Assessment of Reasons for Presentation in New Primary Glaucoma Patients and Identification of Risk Factors for Late Presentation

PURPOSE

To evaluate the reasons for seeking care among South Indian primary glaucoma patients and to determine the relationship of various patient characteristics to glaucoma severity at presentation.

DESIGN

Cross-sectional study.

PARTICIPANTS

One hundred sixty-one new primary open-angle glaucoma (POAG) and primary angle-closure glaucoma (PACG) patients seeking treatment at a tertiary eye hospital.

METHODS

After confirmation of diagnosis, participants' clinical information and their reported reasons for presentation were assessed. Data collected include age, gender, education, occupation, rural or urban residence, distance traveled to the hospital, method of transportation, need for an accompanying person, place of screening before referral or whether they came by themselves for testing. Advanced glaucoma was defined by a cup-to-disc ratio of 0.85 or higher in either eye.

MAIN OUTCOME MEASURES

The primary outcomes were the various reasons for presentation. The secondary outcome was to determine whether a relationship existed between the patient characteristics mentioned above and presentation with advanced glaucoma.

RESULTS

The mean age of the participants was 60.8 years. The primary reason for presentation was defective vision (55.2%) followed by routine ophthalmic evaluation (13%). Sixty-four patients (39.8%) showed advanced glaucoma in at least 1 eye at presentation. Unilateral blindness was noted in 18 patients (11.2%). The mean vertical cup-to-disc ratio was 0.66 (standard deviation [SD], ±0.16) in the better-seeing eye and 0.76 (SD, ±0.12) in the worse-seeing eye. The mean presenting intraocular pressure was 22.9 mmHg in POAG and 25.5 mmHg in PACG patients. Multivariate logistic regression analyses showed that people who are currently unemployed (P < 0.001; odds ratio [OR], 4.19; 95% confidence interval [CI], 1.95-8.99) and rural residence (P = 0.04; OR, 0.46; 95% CI, 0.21-0.99) had greater odds of demonstrating advanced glaucoma at presentation. Presentation with defective vision, older age, and education less than college graduation were associated with greater odds of showing advanced glaucoma in univariate analysis, but not in multivariate analyses.

CONCLUSIONS

In a South Indian population, absence of work and rural residence was associated with advanced glaucoma at presentation. The population in whom these risk factors are common should be targeted for screening and outreach.