Cross-sectional study: Does combining optical coherence tomography measurements using the ‘Random Forest’ decision tree classifier improve the prediction of the presence of perimetric deterioration in glaucoma suspects?

Abstract

ObjectivesTo develop a classifier to predict the presence of visual field (VF) deterioration in glaucoma suspects based on optical coherence tomography (OCT) measurements using the machine learning method known as the ‘Random Forest’ algorithm.DesignCase–control study.Participants293 eyes of 179 participants with open angle glaucoma (OAG) or suspected OAG.InterventionsSpectral domain OCT (Topcon 3D OCT-2000) and perimetry (Humphrey Field Analyser, 24-2 or 30-2 SITA standard) measurements were conducted in all of the participants. VF damage (Ocular Hypertension Treatment Study criteria (2002)) was used as a ‘gold-standard’ to classify glaucomatous eyes. The ‘Random Forest’ method was then used to analyse the relationship between the presence/absence of glaucomatous VF damage and the following variables: age, gender, right or left eye, axial length plus 237 different OCT measurements.Main outcome measuresThe area under the receiver operating characteristic curve (AROC) was then derived using the probability of glaucoma as suggested by the proportion of votes in the Random Forest classifier. For comparison, five AROCs were derived based on: (1) macular retinal nerve fibre layer (m-RNFL) alone; (2) circumpapillary (cp-RNFL) alone; (3) ganglion cell layer and inner plexiform layer (GCL+IPL) alone; (4) rim area alone and (5) a decision tree method using the same variables as the Random Forest algorithm.ResultsThe AROC from the combined Random Forest classifier (0.90) was significantly larger than the AROCs based on individual measurements of m-RNFL (0.86), cp-RNFL (0.77), GCL+IPL (0.80), rim area (0.78) and the decision tree method (0.75; p<0.05).ConclusionsEvaluating OCT measurements using the Random Forest method provides an accurate prediction of the presence of perimetric deterioration in glaucoma suspects.

Fundus image classification methods for the detection of glaucoma: A review

“…Statistical distributions of classification parameters were obtained compared using the Friedman's non‐parametric test (Atto, Pastor, & Mercier, ). Decision tree classifier (Sugimoto et al, ) is a method used for recognizing pattern having if‐then‐else rules of a finite set and represented in a tree‐like structure (Strobl, Malley, & Tutz, ; Sugimoto et al, ) The most common drawback that impacts the diagnostic accuracy of this method is the problem of “overfitting” (Narooie‐Nejad et al, ). Indeed, in comparison with Random Forest method discriminating between pre‐parametric and parametric glaucoma eyes, the decision tree method performed less accurately (Lisboa et al, ).…”

Risk factors of hyperuricemia calculated by random forest machine learning

“…Fourth, the models trained by RF can maintain accuracy even if there are many missing values. Last but not the least, any interaction or correlation between variables does not adversely affect the RF classi cation since it is capable of representing high order interactions [15]. The analysis was accomplished by using Python.…”

Validating the Usefulness of the “Random Forests” Classifier to Diagnose Early Glaucoma With Optical Coherence Tomography

“…We have recently demonstrated the usefulness of a RF classifier to diagnose glaucoma using OCT measurements; 36,37 however, a limitation of our prior studies was the lack of external validation in an independent dataset, instead leave-one-out cross-validation was used to investigate diagnostic performance. Further, the study population consisted of wide range of glaucoma patients with early to advanced-stage disease, but the merit of using OCT to diagnose glaucoma may lie in early stage detection.…”