Combined Evaluation of Frequency Doubling Technology Perimetry and Scanning Laser Ophthalmoscopy for Glaucoma Detection Using Automated Classification

Horn, Folkert K.; Laemmer, Robert; Mardin, Christian Y.; Jünemann, Anselm; Michelson, Georg; Lausen, Berthold; Adler, Werner

doi:10.1097/ijg.0b013e3182027766

Cited by 14 publications

(15 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Random forests have already demonstrated their ability to yield good results for glaucoma classification based on topography information of the eye background [25,26]. In this work, we were interested in the reduction of the computational cost of these classifiers by the application of pruning strategies.…”

Section: Discussionmentioning

confidence: 99%

“…The set of variables comprises 102 features from the Heidelberg Retina Tomograph which produces three dimensional laser scanning images of the eye background and then calculates topographical features of these images. The data set is a combination of two case-control studies performed at the Erlangen University Eye Hospital described in more detail in Horn et al [26] to examine the predictive power for glaucoma detection. In contrast to Horn et al we used only those glaucomatous observations that were part of the test data in [26].…”

Section: Clinical Data Set: Glaucoma Classification With Hrt Measuremmentioning

confidence: 99%

“…The data set is a combination of two case-control studies performed at the Erlangen University Eye Hospital described in more detail in Horn et al [26] to examine the predictive power for glaucoma detection. In contrast to Horn et al we used only those glaucomatous observations that were part of the test data in [26]. Thus, the glaucoma prevalence in our data set was reduced, motivated by the low Figure 1 Classification results obtained with the glaucoma data set: AUC, Brier score on the total data set (top row, from left to right), Brier score in the majority class, and Brier score in the minority class (bottom row, from left to right) for ensembles consisting of 1 to 200 trees where the trees are selected based on the three examined strategies DF pruning, Brier pruning, and UWA pruning and performing random selection.…”

Section: Clinical Data Set: Glaucoma Classification With Hrt Measuremmentioning

confidence: 99%

See 2 more Smart Citations

Ensemble Pruning for Glaucoma Detection in an Unbalanced Data Set

Adler¹,

Gefeller²,

Gul³

et al. 2016

Methods Inf Med

Self Cite

View full text Add to dashboard Cite

SummaryBackground: Random forests are successful classifier ensemble methods consisting of typically 100 to 1000 classification trees. Ensemble pruning techniques reduce the computational cost, especially the memory demand, of random forests by reducing the number of trees without relevant loss of performance or even with increased performance of the sub-ensemble. The application to the problem of an early detection of glaucoma, a severe eye disease with low prevalence, based on topographical measurements of the eye background faces specific challenges. Objectives:We examine the performance of ensemble pruning strategies for glaucoma detection in an unbalanced data situation. Methods:The data set consists of 102 topographical features of the eye background of 254 healthy controls and 55 glaucoma patients. We compare the area under the receiver operating characteristic curve (AUC), and the Brier score on the total data set, in the majority class, and in the minority class of pruned random forest ensembles obtained with strategies based on the prediction accuracy of greedily grown sub-ensembles, the uncertainty weighted accuracy, and the similarity between single trees. To validate the findings and to examine the influence of the prevalence of glaucoma in the data set, we additionally perform a simulation study with lower prevalences of glaucoma. Results:In glaucoma classification all three pruning strategies lead to improved AUC and smaller Brier scores on the total data set with sub-ensembles as small as 30 to 80 trees compared to the classification results obtained with the full ensemble consisting of 1000 trees. In the simulation study, we were able to show that the prevalence of glaucoma is a critical factor and lower prevalence decreases the performance of our pruning strategies. Conclusion:The memory demand for glaucoma classification in an unbalanced data situation based on random forests could effectively be reduced by the application of pruning strategies without loss of performance in a population with increased risk of glaucoma.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Clinical Data Set: Glaucoma Classification With Hrt Measuremmentioning

confidence: 99%

Section: Clinical Data Set: Glaucoma Classification With Hrt Measuremmentioning

confidence: 99%

See 1 more Smart Citation

Ensemble Pruning for Glaucoma Detection in an Unbalanced Data Set

Adler¹,

Gefeller²,

Gul³

et al. 2016

Methods Inf Med

Self Cite

View full text Add to dashboard Cite

show abstract

“…Since its inception, MLCs have been studied in combination with several apparatus designed to improve the diagnosis of glaucoma such as TD-OCT [11, 15, 17, 19], SD-OCT [21, 22], Heidelberg Retina Tomograph (HRT) [16, 18, 24], Scanning Laser Polarimetry (GDx) [14], and VF [12, 17, 18, 20, 22]. …”

Section: Discussionmentioning

confidence: 99%

“…MLCs train computerized systems to detect the relationship between multiple input parameters, eventually facilitating the diagnosis of a condition. In fact, some reports suggest that MLCs are as good as [10–13] or better than [14–20] currently available techniques for glaucoma diagnosis.…”

Section: Introductionmentioning

confidence: 99%

Glaucoma Diagnostic Accuracy of Machine Learning Classifiers Using Retinal Nerve Fiber Layer and Optic Nerve Data from SD-OCT

Barella

Costa

Vidotti

et al. 2013

Journal of Ophthalmology

View full text Add to dashboard Cite

Purpose. To investigate the diagnostic accuracy of machine learning classifiers (MLCs) using retinal nerve fiber layer (RNFL) and optic nerve (ON) parameters obtained with spectral domain optical coherence tomography (SD-OCT). Methods. Fifty-seven patients with early to moderate primary open angle glaucoma and 46 healthy patients were recruited. All 103 patients underwent a complete ophthalmological examination, achromatic standard automated perimetry, and imaging with SD-OCT. Receiver operating characteristic (ROC) curves were built for RNFL and ON parameters. Ten MLCs were tested. Areas under ROC curves (aROCs) obtained for each SD-OCT parameter and MLC were compared. Results. The mean age was 56.5 ± 8.9 years for healthy individuals and 59.9 ± 9.0 years for glaucoma patients (P = 0.054). Mean deviation values were −1.4 dB for healthy individuals and −4.0 dB for glaucoma patients (P < 0.001). SD-OCT parameters with the greatest aROCs were cup/disc area ratio (0.846) and average cup/disc (0.843). aROCs obtained with classifiers varied from 0.687 (CTREE) to 0.877 (RAN). The aROC obtained with RAN (0.877) was not significantly different from the aROC obtained with the best single SD-OCT parameter (0.846) (P = 0.542). Conclusion. MLCs showed good accuracy but did not improve the sensitivity and specificity of SD-OCT for the diagnosis of glaucoma.

show abstract

Multimodal Screening of Glaucoma Improves Sensitivity and Specificity

Horn

Adler

2014

Teleophthalmology in Preventive Medicine

View full text Add to dashboard Cite

Combined Evaluation of Frequency Doubling Technology Perimetry and Scanning Laser Ophthalmoscopy for Glaucoma Detection Using Automated Classification

Cited by 14 publications

References 27 publications

Ensemble Pruning for Glaucoma Detection in an Unbalanced Data Set

Ensemble Pruning for Glaucoma Detection in an Unbalanced Data Set

Glaucoma Diagnostic Accuracy of Machine Learning Classifiers Using Retinal Nerve Fiber Layer and Optic Nerve Data from SD-OCT

Multimodal Screening of Glaucoma Improves Sensitivity and Specificity

Contact Info

Product

Resources

About