EffUnet-SpaGen: An Efficient and Spatial Generative Approach to Glaucoma Detection
Autor: | Colin E. Willoughby, Srinivasan Kavitha, Bryan M. Williams, Rengaraj Venkatesh, Gabriela Czanner, David S. Friedman, Silvester Czanner, Venkatesh Krishna Adithya |
---|---|
Jazyk: | angličtina |
Rok vydání: | 2021 |
Předmět: |
Computer science
diagnosis Computer applications to medicine. Medical informatics R858-859.7 Article 030218 nuclear medicine & medical imaging Convolution 03 medical and health sciences 0302 clinical medicine Photography Medical imaging Redundancy (engineering) medicine Radiology Nuclear Medicine and imaging Segmentation generative model Electrical and Electronic Engineering TR1-1050 Block (data storage) business.industry Pattern recognition QA75.5-76.95 Computer Graphics and Computer-Aided Design Generative model medicine.anatomical_structure glaucoma machine learning classification Electronic computers. Computer science 030221 ophthalmology & optometry Benchmark (computing) Computer Vision and Pattern Recognition Artificial intelligence business Optic disc |
Zdroj: | Journal of Imaging Volume 7 Issue 6 Journal of Imaging, Vol 7, Iss 92, p 92 (2021) |
ISSN: | 2313-433X |
DOI: | 10.3390/jimaging7060092 |
Popis: | Current research in automated disease detection focuses on making algorithms “slimmer” reducing the need for large training datasets and accelerating recalibration for new data while achieving high accuracy. The development of slimmer models has become a hot research topic in medical imaging. In this work, we develop a two-phase model for glaucoma detection, identifying and exploiting a redundancy in fundus image data relating particularly to the geometry. We propose a novel algorithm for the cup and disc segmentation “EffUnet” with an efficient convolution block and combine this with an extended spatial generative approach for geometry modelling and classification, termed “SpaGen” We demonstrate the high accuracy achievable by EffUnet in detecting the optic disc and cup boundaries and show how our algorithm can be quickly trained with new data by recalibrating the EffUnet layer only. Our resulting glaucoma detection algorithm, “EffUnet-SpaGen”, is optimized to significantly reduce the computational burden while at the same time surpassing the current state-of-art in glaucoma detection algorithms with AUROC 0.997 and 0.969 in the benchmark online datasets ORIGA and DRISHTI, respectively. Our algorithm also allows deformed areas of the optic rim to be displayed and investigated, providing explainability, which is crucial to successful adoption and implementation in clinical settings. |
Databáze: | OpenAIRE |
Externí odkaz: |