June 2021
Volume 62, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2021
The effect of dataset size on pretraining and transfer learning for diabetic retinopathy in narrow field to widefield domain shift
Author Affiliations & Notes
  • Mhd Hasan Sarhan
    Carl Zeiss Meditec AG, Munich, Germany
    Technische Universitat Munchen, Munich, Bayern, Germany
  • Niranchana Manivannan
    Carl Zeiss Meditec Inc, Dublin, California, United States
  • Mary K Durbin
    Carl Zeiss Meditec Inc, Dublin, California, United States
  • Susan Su
    Carl Zeiss Meditec Inc, Dublin, California, United States
  • Angelina Covita
    Carl Zeiss Meditec Inc, Dublin, California, United States
  • Abouzar Eslami
    Carl Zeiss Meditec AG, Munich, Germany
  • Footnotes
    Commercial Relationships   Mhd Hasan Sarhan, Carl Zeiss Meditec, AG (E); Niranchana Manivannan, Carl Zeiss Meditec, Inc. (E); Mary Durbin, Carl Zeiss Meditec, Inc. (E); Susan Su, Carl Zeiss Meditec, Inc. (C); Angelina Covita, Carl Zeiss Meditec, Inc. (E); Abouzar Eslami, Carl Zeiss Meditec, AG (E)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2021, Vol.62, 2143. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Mhd Hasan Sarhan, Niranchana Manivannan, Mary K Durbin, Susan Su, Angelina Covita, Abouzar Eslami; The effect of dataset size on pretraining and transfer learning for diabetic retinopathy in narrow field to widefield domain shift. Invest. Ophthalmol. Vis. Sci. 2021;62(8):2143.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : To study the effect of the pretraining dataset on the transferred model performance. It has been established that pretraining a deep learning system (DLS) on narrow field (NF) funds images then transfer the model to widefield (WF) fundus images enhances the performance on WF fundus diagnosis for diabetic retinopathy (DR). We studied the effect of the pretraining dataset size on the transferred model.

Methods : We collected smaller (32k) and larger (120k) NF images datasets using VISUSCOUT® 100 (ZEISS, Jena, Germany) handheld fundus camera. The images were annotated for referable DR (more than mild DR). A DLS (Resnet-50) for each dataset was developed to detect referable DR from the images. For the transfer learning, we collected WF fundus images from 361 subjects (77% normal, 23% referable DR) using tabletop CLARUSTM 500 (ZEISS, Dublin, CA). The images were annotated for referable DR. We split the WF dataset into train/validation/test (80/10/10) sets and evaluated the final model on the test set. The two NF models were fine-tuned on the same training split of WF data and compared together and with directly evaluating WF images on NF models without fine-tuning. Augmentation techniques were used for robustness. We compared accuracy, sensitivity (Sen), specificity (Spc), and area under the curve of the receiver operating characteristics curve (AUC) for evaluation of the performance. Fig. 1 overviews the performed experiments

Results : The pretrained model on the larger NF dataset produced marginally better results on the WF dataset after transfer compared with the pretrained model on the smaller NF dataset on the considered metrics. We observed large confidence intervals for Sen and Spc because of the dataset size limitations. The best model achieved 86.5% AUC, 84.4% accuracy, 90% Sen (95%CI [71%, 100%]), and 83% Spc (95%CI [70%, 95%])

Conclusions : Increasing the NF pretraining dataset size potentially enhances the fine-tuning on the WF dataset. However, this enhancement is limited by the size of the WF dataset. This suggests that increasing the NF pretraining dataset size improves the prediction on WF dataset up to a certain limit where collecting more WF data is required for further improvements.

This is a 2021 ARVO Annual Meeting abstract.

 

Overview of the performed experiments

Overview of the performed experiments

 

Evaluation results on the test set of WF dataset. The pretraining is done on the NF datasets and the fine-tuning results are compared

Evaluation results on the test set of WF dataset. The pretraining is done on the NF datasets and the fine-tuning results are compared

×
×

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.

×