MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641054709 A) filed by Manipal Academy Of Higher Education on April 29, 2026, for System And Method For Foveal Region Guided Convolutional Feature Extraction From Optical Coherence Tomography Images.
Inventors include Pavithra K C; Preetham Kumar; Geetha M; and Sulatha V Bhandary.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The invention provides a system (100) comprising a data collection module (301), data annotation module (303), foveal region localisation and extraction module (305), data processing module (307), data management and partitioning module (309), classification processing module (311), data evaluation module (313), feature attribution and visualisation module (315), and output interface (317). The foveal region localisation and extraction module (305) applies a calibration factor of 156 pixels per millimetre extracting a 156×625-pixel foveal region from each 938×625-pixel optical coherence tomography image, reducing input dimensionality to 16.6%. The classification processing module (311) achieves mean accuracy of 0.9854, area under the receiver operating characteristic curve of 0.9952, and Matthews Correlation Coefficient of 0.9712 with 70% reduced peak graphics processing unit memory and 47% reduced inference latency relative to whole-image processing. FIG.1C
Disclaimer: Curated by HT Syndication.
