Automatic Low-quality Fundus Image Enhancement and Diabetic Retinopathy Grading with Explanations<
Osmar Zaiane
March 22, 2024, Friday, 2:00 PM - 3:00 PM EDT
Abstract
We present two lines of work that are connected but not yet put together, namely the automatic enhancement of retinal image quality and the classification of retinal images.

Retinal fundus images have been applied for the diagnosis and screening of eye diseases, such as Diabetic Retinopathy or Diabetic Macular Edema. However, low-quality fundus images potentially increase uncertainty in the diagnosis of eye fundus disease and even lead to misdiagnosis by ophthalmologists. We explore the potential of a self-supervised task for improving the quality of fundus images without the requirement of high-quality reference images.

Traditional Diabetic Retinopathy (DR) automatic classification algorithms rely on the precise detection of microaneurysms and hemorrhage lesions. Such lesion annotation is an expensive and time-consuming process and therefore it is expected to develop automatic grading methods with only image-level annotations. We formulate the weakly supervised DR grading as a multi-instance learning problem and propose a domain adaptation multi-instance learning with attention mechanism for DR grading.
Bio
Osmar R. Zaiane is a Professor in Computing Science at the University of Alberta, Canada, Fellow of the Alberta Machine Intelligence Institute (Amii), and Canada CIFAR AI Chair. He is also a Fellow of the Canadian Academy of Engineering. Dr. Zaiane obtained his Ph.D. from Simon Fraser University, Canada, in 1999. He has published more than 400 papers in refereed international conferences and journals. He is Associate Editor of many International Journals on data mining and data analytics and served as program chair and general chair for scores of international conferences in the field of knowledge discovery and data mining. Dr. Zaiane received numerous awards including the Killam Professorship award, the McCalla Research Professorship, and the ACM SIGKDD Service Award from the ACM Special Interest Group on Data Mining, which runs the world's premier data science, big data, and data mining association and conference.