Automated HBsAg Stain Analysis for Accurate Liver Diagnosis
Automated HBsAg Stain Analysis for Accurate Liver Diagnosis
Quantitative Analysis of Immunohistochemical Staining of Liver Slides
Tags: National Cheng Kung University, Taiwan, Healthcare & Lifesciences
This technology automates the quantification of Hepatitis B surface antigen (HBsAg) intensity in immunohistochemically stained whole slide images (WSI) using a convolutional neural network combined with image processing. It classifies each pixel's staining intensity from 0+ to 3+ and has shown strong detection and classification performance. The method provides pathologists with objective quantitative data, improving diagnostic accuracy for liver diseases and potentially aiding in recurrence prediction of liver cancer. Its applications extend to automated diagnostics in clinical pathology and the development of innovative medical technologies. The method reduces the manual burden of visual analysis, enhancing efficiency and consistency in liver pathology diagnosis.
IP Type or Form Factor: Patent Granted; Software & Algorithm
TRL: 6 - prototype validated in intended environment
Industry or Tech Area: Diagnostics & Screening; Biotechnology
