MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641079056 A) filed by Sr University on June 25, 2026, for Early Detection Of Kidney Tumor Consequences Using Image-Based Fine- Grained Refinement Learning Algorithm (i-Fgrl).
Inventors include A. Deepika; and Dr. Santosh Kumar Henge.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: EARLY DETECTION OF KIDNEY TUMOR CONSEQUENCES USING IMAGE-BASED FINE-GRAINED REFINEMENT LEARNING ALGORITHM (I-FGRL) A computer-implemented Image-based Fine-Grained Refinement Learning Algorithm (I- FGRL) for early kidney tumor detection is disclosed. The system comprises a preprocessing module, a Fine-Grained Layer, a Refinement Layer, and a Fusion Layer configured to analyze kidney CT scan images. A Tumor-Aware Noise Reduction (TANR) mechanism utilizes tumor probability maps to perform adaptive denoising while preserving tumor boundaries, edge contrast, and textural information. The Fine-Grained Layer extracts structural and morphological tumor features, the Refinement Layer suppresses noise and enhances tumor-specific characteristics, and the Fusion Layer combines refined local features with global contextual information for accurate tumor localization and classification. The integrated architecture progressively enhances micro-tumor regions, reduces false-positive detections, improves classification accuracy, and provides reliable clinical decision support for early diagnosis and treatment planning of kidney tumors
Disclaimer: Curated by HT Syndication.
