数字病理图像智能预测免疫组化软件（IHCPred） The WIFPS was built on WSIs for improving the subtyping of lung cancer from surgical resection and biopsy specimens, which consisted of one binary histological classification model (HCM) to recognize tumor versus normal and one immunohistochemical feature prediction model (IPM) to predict IHC expression status directly from WSIs.

WSI-based immunohistochemical feature prediction system

A total of 1855 WSIs from 1855 individuals of lung specimens obtained from three independent hospitals in China were used for a WSI-based immunohistochemical feature prediction system (WIFPS) development and validation. The WIFPS was built on WSIs for improving the subtyping of lung cancer from surgical resection and biopsy specimens, which consisted of one binary histological classification model (HCM) to recognize tumor versus normal and one immunohistochemical feature prediction model (IPM) to predict IHC expression status directly from WSIs. Its performance was evaluated using surgical resection specimens from one internal validation cohort and one external cohort and biopsy specimens from two internal validation cohorts and one external validation cohort.