SPLICE -- Streamlining Digital Pathology Image Processing

• DOI: 10.48550/arxiv.2404.17704
• Publication Date: 2024-04-26
• Authors: Areej Alsaafin; Peyman Nejat; Abubakr Shafique; Jibran Khan; Saghir Alfasly; Ghazal Alabtah; Hamid R. Tizhoosh
• Keywords: splice; Digital Pathology; Digital Image Analysis

Digital pathology and the integration of artificial intelligence (AI) models have revolutionized histopathology, opening new opportunities. With increasing availability Whole Slide Images (WSIs), there's a growing demand for efficient retrieval, processing, analysis relevant images from vast biomedical archives. However, processing WSIs presents challenges due to their large size content complexity. Full computer digestion is impractical, all patches individually prohibitively expensive. In this paper, we propose an unsupervised patching algorithm, Sequential Patching Lattice Image Classification Enquiry (SPLICE). This novel approach condenses histopathology WSI into compact set representative patches, forming "collage" while minimizing redundancy. SPLICE prioritizes patch quality uniqueness by sequentially analyzing selecting non-redundant features. We evaluated search match applications, demonstrating improved accuracy, reduced computation time, storage requirements compared existing state-of-the-art methods. As method, effectively reduces representing tissue 50%. reduction enables numerous algorithms in computational operate much more efficiently, paving way accelerated adoption digital pathology.