Article Text

Download PDFPDF
Anterior segment biometric measurements explain misclassifications by a deep learning classifier for detecting gonioscopic angle closure
  1. Alice Shen1,
  2. Michael Chiang1,
  3. Anmol A Pardeshi1,
  4. Roberta McKean-Cowdin2,
  5. Rohit Varma3,
  6. Benjamin Y Xu1
  1. 1Department of Ophthalmology, USC Keck School of Medicine, Los Angeles, California, USA
  2. 2Department of Preventive Medicine, USC Keck School of Medicine, Los Angeles, California, USA
  3. 3Southern California Eye Institute, CHA Hollywood Presbyterian Medical Center, Los Angeles, California, USA
  1. Correspondence to Dr Benjamin Y Xu, Department of Ophthalmology, USC Keck School of Medicine, Los Angeles, CA 90033, USA; Benjamin.Xu{at}med.usc.edu

Abstract

Background/aims To identify biometric parameters that explain misclassifications by a deep learning classifier for detecting gonioscopic angle closure in anterior segment optical coherence tomography (AS-OCT) images.

Methods Chinese American Eye Study (CHES) participants underwent gonioscopy and AS-OCT of each angle quadrant. A subset of CHES AS-OCT images were analysed using a deep learning classifier to detect positive angle closure based on manual gonioscopy by a reference human examiner. Parameter measurements were compared between four prediction classes: true positives (TPs), true negatives (TNs), false positives (FPs) and false negatives (FN). Logistic regression models were developed to differentiate between true and false predictions. Performance was assessed using area under the receiver operating curve (AUC) and classifier accuracy metrics.

Results 584 images from 127 participants were analysed, yielding 271 TPs, 224 TNs, 77 FPs and 12 FNs. Parameter measurements differed (p<0.001) between prediction classes among anterior segment parameters, including iris curvature (IC) and lens vault (LV), and angle parameters, including angle opening distance (AOD). FP resembled TP more than FN and TN in terms of anterior segment parameters (steeper IC and higher LV), but resembled TN more than TP and FN in terms of angle parameters (wider AOD). Models for detecting FP (AUC=0.752) and FN (AUC=0.838) improved classifier accuracy from 84.8% to 89.0%.

Conclusions Misclassifications by an OCT-based deep learning classifier for detecting gonioscopic angle closure are explained by disagreement between anterior segment and angle parameters. This finding could be used to improve classifier performance and highlights differences between gonioscopic and AS-OCT definitions of angle closure.

  • angle
  • diagnostic tests/investigation
  • glaucoma
  • imaging

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author (BYX), upon reasonable request.

Statistics from Altmetric.com

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author (BYX), upon reasonable request.

View Full Text

Footnotes

  • Twitter @BenXuLab

  • Contributors AS, RM-C, RV and BYX conceived and designed this study. AS, MC, AAP, RM-C, RV and BYX contributed to the drafting of manuscript. AS, MC, AAP and BYX contributed to the data analysis and interpretation. All authors contributed to the critical appraisal and final approval of the manuscript. BYX provided the overall supervision of this work.

  • Funding This work was supported by grants U10 EY017337, P30 EY029220 and K23 EY029763 from the National Eye Institute, National Institute of Health, Bethesda, Maryland; a Young Clinician Scientist Research Award from the American Glaucoma Society (no grant number), San Francisco, California; a Grant-in-Aid Research Award from Fight for Sight (no grant number), New York, New York; a SC-CTSI Clinical and Community Research Award from the Southern California Clinical and Translational Science Institute (no grant number), Los Angeles, California; and an unrestricted grant to the Department of Ophthalmology from Research to Prevent Blindness (no grant number), New York, New York.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.