June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
The suitability of color histogram-based features for keratoconus detection from corneal thickness with and neural networks.
Rossen M Hazarbassanov; Laith Al-Zubaidi; Zahraa M Mosa; Hazem Abdul Mutal; Alexandru Lavric; Hidenori Takahashi; Taneri Taneri; Siamak Yousefi; Ali Al-Timemy
Author Affiliations & Notes
  • Rossen M Hazarbassanov
    Ophthalmology and Visual Sciences, Universidade Federal de Sao Paulo Escola Paulista de Medicina, Sao Paulo, SP, Brazil
  • Laith Al-Zubaidi
    School of Mechanical, Medical and Process Engineering, Queensland University of Technology, Brisbane, Queensland, Australia
  • Zahraa M Mosa
    Pharmacy, Uruk University, Baghdad, Iraq
  • Hazem Abdul Mutal
    Ophthalmology, Assiut University Faculty of Medicine, Assiut, Egypt
  • Alexandru Lavric
    Computers, Electronics and Automation, Stefan cel Mare University of Suceava, Suceava, Romania
  • Hidenori Takahashi
    Ophthalmology, Jichi Ika Daigaku, Shimotsuke, Tochigi, Japan
  • Taneri Taneri
    University Eye-Clinic, Ruhr-Universitat Bochum, Bochum, Nordrhein-Westfalen, Germany
    Zentrum für Refraktive Chirurgie, Muenster, Germany
  • Siamak Yousefi
    Genetics, Genomics, and Informatics, The University of Tennessee Health Science Center VolShop Memphis, Memphis, Tennessee, United States
    Ophthalmology, The University of Tennessee Health Science Center VolShop Memphis, Memphis, Tennessee, United States
  • Ali Al-Timemy
    Biomedical Engineering Department, AL-Khwarizmi College of Engineering, University of Baghdad, Baghdad, Baghdad , Iraq
    Computing and Mathematics, University of Plymouth, Plymouth, United Kingdom
  • Footnotes
    Commercial Relationships   Rossen Hazarbassanov None; Laith Al-Zubaidi None; Zahraa Mosa None; Hazem Mutal None; Alexandru Lavric None; Hidenori Takahashi None; Taneri Taneri None; Siamak Yousefi None; Ali Al-Timemy None
  • Footnotes
    Support  Ali H. Al-Timemy acknowledges the ARVO Collaborative Research Fellowship
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 1089. doi:
  • Views
  • Share
  • Tools
    • Alerts
      User Alerts
      You are adding an alert for:
      The suitability of color histogram-based features for keratoconus detection from corneal thickness with and neural networks.
      You will receive an email whenever this article is corrected, updated, or cited in the literature. You can manage this and all other alerts in My Account
      The alert will be sent to:
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation
      Citation

      Rossen M Hazarbassanov, Laith Al-Zubaidi, Zahraa M Mosa, Hazem Abdul Mutal, Alexandru Lavric, Hidenori Takahashi, Taneri Taneri, Siamak Yousefi, Ali Al-Timemy; The suitability of color histogram-based features for keratoconus detection from corneal thickness with and neural networks.. Invest. Ophthalmol. Vis. Sci. 2023;64(8):1089.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

      ×
    • Get Permissions
  • Supplements
Abstract

Purpose : To develop and validate a neural network model for detecting keratoconus (KCN) based on color histogram features of corneal thickness map.

Methods : A total of 871 corneal thickness maps were collected from 871 eyes using Pentacam instruments (Oculus Optikgerate GmbH) in an eye clinic in Egypt. We developed a neural network framework based on color histogram features of corneal thickness map to detect KCN with 10-fold cross validation. We preprocessed the maps then generated three histograms corresponding to red, green, and blue channels followed by histogram Mean, Variance, Skewness, Kurtosis, Energy, and Entropy. We then evaluated the performance of the model based on accuracy, precision and F-score and area under the receiver operating characteristic curve (AUC) metrics (Fig.1).

Results : The development dataset included 377 normal eyes and 289 eyes with KCN that were previously diagnosed by two corneal specialists. The independent validation data set consisted of 123 normal eyes and 82 eyes with KCN. The accuracy and AUC of the proposed model in detecting KCN from corneal thickness maps based on the independent validation dataset were 97.1% and 0.98, respectively. The precision and F-score of the model were 0.97 and 0.971, respectively.

Conclusions : We developed a simple machine learning model based on the color histogram features to detect KCN and evaluated it based on corneal thickness map. While the model was developed based on relatively a small number of samples, it was accurate and generalizable. This approach is more appropriate for the scenarios with small subsets and may eventually improve KCN research and clinical practice.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Figure 1. Block diagram of the proposed approach based on color histogram features and neural networks applied to the corneal thickness map
View OriginalDownload Slide

Figure 1. Block diagram of the proposed approach based on color histogram features and neural networks applied to the corneal thickness map

Figure 1. Block diagram of the proposed approach based on color histogram features and neural networks applied to the corneal thickness map

Figure 1. Block diagram of the proposed approach based on color histogram features and neural networks applied to the corneal thickness map
View OriginalDownload Slide

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Attribution-NonCommercial-NoDerivatives
Copyright © 2024 Association for Research in Vision and Ophthalmology.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×
This site uses cookies. By continuing to use our website, you are agreeing to our privacy policy.Accept