EyeWeS: Weakly Supervised Pre-Trained Convolutional Neural Networks for Diabetic Retinopathy Detection

• Type: Article in International Conference Proceedings Book
• Date: 2019

Authors

Pedro Costa (Author) Other; Teresa Araújo (Author) Other; Guilherme Aresta (Author) Other; Adrian Galdran (Author) Other; Ana Maria Mendonça (Author) FEUP; Asim Smailagic (Author) Other; Aurélio Campilho (Author) FEUP

Conference proceedings International

Title: PROCEEDINGS OF MVA 2019 16TH INTERNATIONAL CONFERENCE ON MACHINE VISION APPLICATIONS (MVA)

Pages: 1-6

16th International Conference on Machine Vision Applications (MVA)

Tokyo, JAPAN, MAY 27-31, 2019

Indexing

ISI Web of Knowledge - 2 Citations

Scopus - 12 Citations

Other information

Authenticus ID: P-00Q-TNA

DOI: 10.23919/mva.2019.8757991

Abstract (EN): Diabetic Retinopathy (DR) is one of the leading causes of preventable blindness in the developed world. With the increasing number of diabetic patients there is a growing need of an automated system for DR detection. We propose EyeWeS, a method that not only detects DR in eye fundus images but also pinpoints the regions of the image that contain lesions, while being trained with image labels only. We show that it is possible to convert any pre-trained convolutional neural network into a weakly-supervised model while increasing their performance and efficiency. EyeWeS improved the results of Inception V3 from 94:9% Area Under the Receiver Operating Curve (AUC) to 95:8% AUC while maintaining only approximately 5% of the Inception V3's number of parameters. The same model is able to achieve 97:1% AUC in a cross-dataset experiment.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 6