Fig. 1 | Scientific Data

Fig. 1

From: PathoGraph: A Graph-Based Method for Standardized Representation of Pathology Knowledge

Fig. 1

Pathology knowledge contained in different modalities of pathology data. (a) The compositions of the diseased tissue and phenotypes in a thyroid carcinoma hematoxylin and eosin-stained (HE) slide, reflecting the characteristics of the disease at the tissue and cellular level. (b) Two lymphoma immunohistochemistry (IHC) slides stained for the antigens of CD20 and OCT2, respectively. The brown areas in the slides are the stained antigens, where CD20 are stained in the tumor cell membranes (i.e. CD20 resides in the membranes) and OCT2 are stained in the nuclei. The normal cells, which do not stain for these antigens, appear blue. The immunophenotypes of CD20 and OCT2 reflect the characteristics of the disease at the protein level. (c) The schematic representation of a pathologist's diagnostic process to subtype a patient with lymphoma obtained from a pathology report. Firstly, the phenotypes in the HE slide lead to an initial diagnosis that this tumor is a lymphoma. Secondly, the pathologist considers three possible subtypes of lymphoma aligning with these phenotypes. To differentiate these diagnostic possibilities, additional IHC results are introduced, such as CD3, CD5, etc. The IHC results support large B-cell lymphoma and contradict the other subtypes, leading to its establishment as the final diagnosis.

Back to article page