Litcius/Paper detail

Neural network application for semantic segmentation of fundus

Rustam Paringer, IPSI RAS – Branch of the FSRC “Crystallography and Photonics” RAS, Artem Mukhin, Nataly Ilyasova, IPSI RAS – Branch of the FSRC “Crystallography and Photonics” RAS, Nikita Demin, IPSI RAS – Branch of the FSRC “Crystallography and Photonics” RAS

2022Computer Optics13 citationsDOIOpen Access PDF

Abstract

Advances in the neural networks have brought revolution in many areas, especially those related to image processing and analysis. The most complex is a task of analyzing biomedical data due to a limited number of samples, imbalanced classes, and low-quality labelling. In this paper, we look into the possibility of using neural networks when solving a task of semantic segmentation of fundus. The applicability of the neural networks is evaluated through a comparison of image segmentation results with those obtained using textural features. The neural networks are found to be more accurate than the textural features both in terms of precision (~25%) and recall (~50%). Neural networks can be applied in biomedical image segmentation in combination with data balancing algorithms and data augmentation techniques.

Topics & Concepts

Computer scienceSegmentationArtificial intelligenceArtificial neural networkTask (project management)Pattern recognition (psychology)Image segmentationFundus (uterus)Image (mathematics)Machine learningEconomicsOphthalmologyManagementMedicineRetinal Imaging and AnalysisBrain Tumor Detection and ClassificationAI in cancer detection
Neural network application for semantic segmentation of fundus | Litcius