Multi-resolution tile-based follicle detection using color and textural information of follicular lymphoma IHC slides

In this work, we introduce a computational approach that automatically detects stained follicles in IHC-stained follicular lymphoma slides using various biomarkers. This novel approach is to process whole-slide and Giga-byte scaled pathology images at multi-resolution levels. The average segmentation accuracy achieves at 83.09±6.25%. Such a computerized analysis of images is expected to provide a better understanding of visual knowledge in digital pathology and thereafter lead to more effective decision support systems for pathologists.