Interpretation of whole-slide images in digital pathology

A computing system for interpreting a whole-slide image of a tissue specimen, the system comprising:

• memory for storing the whole-slide image;
• image processing logic adapted, for at least one region of interest in the tissue specimen, to detect biological entities in that region independently at each of a plurality of magnification levels in the whole-slide image;
• entity graph logic adapted, for each magnification level, to generate an entity graph comprising nodes, representing respective biological entities detected at that magnification level, interconnected by edges representing interactions between entities represented by the nodes;
• hierarchical graph logic adapted, for each region of interest of the at least one region of interest, to generate a hierarchical graph, defining a hierarchy of the entity graphs for that region of interest, in which nodes of different entity graphs are interconnected by hierarchical edges representing hierarchical relations between nodes of the entity graphs for the region of interest;
• and
• an AI system which is adapted to process the hierarchical graph for the at least one region of interest to produce result data corresponding to a medical evaluation of the tissue specimen represented thereby, and to output the result data for the tissue specimen.

A computer program product for interpreting a whole-slide image of a tissue specimen, said computer program product comprising a computer readable storage medium having program instructions embodied therein, the program instructions being executable by a computing system to cause the computing system to:

• for at least one region of interest in the tissue specimen, detect biological entities in that region independently at each of a plurality of magnification levels in the whole-slide image and, for each magnification level, generate an entity graph comprising nodes, representing respective biological entities detected at that magnification level, interconnected by edges representing interactions between entities represented by the nodes;
• for each region of interest, generate a hierarchical graph, defining a hierarchy of the entity graphs for that region of interest, in which nodes of different entity graphs are interconnected by hierarchical edges representing hierarchical relations between nodes of the entity graphs for the region of interest;
• supply the hierarchical graph for said at least one region of interest to an AI system which is adapted to process such hierarchical graphs to produce result data corresponding to a medical evaluation of the tissue specimen represented thereby;
• and
• output the result data for the tissue specimen.