Investigative Ophthalmology & Visual Science Cover Image for Volume 61, Issue 9
July 2020
Volume 61, Issue 9
Free
ARVO Imaging in the Eye Conference Abstract  |   July 2020
The importance of learned vs. task-inspired preprocessing for device generalization capabilities of automated diabetic retinopathy screening algorithms
Author Affiliations & Notes
  • SANDIPAN CHAKROBORTY
    Center for Applications and Research in India, Carl Zeiss India (Bangalore) Pvt. Ltd., Bengaluru, India
  • Alexander Freytag
    Corporate Research and Technology, Carl Zeiss AG, Jena, Germany
  • Ghazal Ghazaei
    Corporate Research and Technology, Carl Zeiss AG, Munich, Germany
  • Anvesh Vankayala
    Center for Applications and Research in India, Carl Zeiss India (Bangalore) Pvt. Ltd., Bengaluru, India
  • Footnotes
    Commercial Relationships   SANDIPAN CHAKROBORTY, Carl Zeiss India (Bangalore) Pvt. Ltd, ZEISS GROUP, Bangalore - 560099, Karnataka, India (E); Alexander Freytag, Carl Zeiss AG, Jena, Germany (E); Ghazal Ghazaei, Carl Zeiss AG, Munich, Germany (E); Anvesh Vankayala, Carl Zeiss India (Bangalore) Pvt. Ltd, ZEISS GROUP, Bangalore - 560099, Karnataka, India (C)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science July 2020, Vol.61, PB0099. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      SANDIPAN CHAKROBORTY, Alexander Freytag, Ghazal Ghazaei, Anvesh Vankayala; The importance of learned vs. task-inspired preprocessing for device generalization capabilities of automated diabetic retinopathy screening algorithms. Invest. Ophthalmol. Vis. Sci. 2020;61(9):PB0099.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Diabetic retinopathy (DR) is one of the most common causes of blindness within diabetic population. Screening using fundus images helps to detect DR at early stages. Various types of fundus cameras are available today, which exhibit stark technical differences, e.g., with respect to image resolution, colorization, illumination, and acquisition quality in general. To screen a large population, it is essential to develop an automated DR screening system that is not tailored to a specific fundus camera device but is instead reliably applicable to multiple device types.

Methods : We compare two complementary approaches: learned preprocessing (LPP) and task-inspired preprocessing (TIPP). For LPP, images are resized to constant width and height and directly presented to a convolutional neural network (CNN). Thereby, the CNN needs to learn adequate preprocessing. As TIPP, we build on the well-known Graham algorithm and modify it such that normalization is applied after disc cropping to avoid device-dependent normalization inconsistencies. Resizing is done to a disc’s normalized radius of 125px. A radius of 15px is used for local mean subtraction as a compromise between speed and visual appearance.

We train InceptionResnet V2 from scratch using Tensorflow. We restrict training data to 133k images recorded with VISUSCOUT® 100 (ZEISS, Jena, Germany), a hand-held low-cost fundus device. As in-device-test, we evaluate trained models on a 15k VISUHEALTH® hold-out test set. As device-generalization-test, we evaluate trained models on E-Optha and Messidor-2 (open source fundus data sets) which were recorded with tabletop cameras.

Results : Prediction accuracy for DR screening is shown in Fig 2. LPP drastically falls short for data from different devices (blue middle and right). In contrast, TIPP barely reduces accuracy on the original device type (orange left), but clearly lifts generalization performance (orange middle and right).

Conclusions : From the perspective of general screening scenarios, we conclude that task-inspired preprocessing may be applied to allow for reliable cross-device predictions.

This is a 2020 Imaging in the Eye Conference abstract.

 

Top: Images of VISUHEALTH ® (left), E-Optha (middle), and Messidor-2 (right). Bottom: the same images after pre-processing

Top: Images of VISUHEALTH ® (left), E-Optha (middle), and Messidor-2 (right). Bottom: the same images after pre-processing

 

DR prediction accuracy (AUC) with learned or task-inspired preprocessing for data from different device types.

DR prediction accuracy (AUC) with learned or task-inspired preprocessing for data from different device types.

×
×

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.

×