Method and device for processing pathological slide image

Inventors

Kang, Jeong Seok • Yoo, Dong Geun • Cho, Soo Ick • Jung, Won Kyung

Assignees

Lunit Inc

Abstract

A computing device includes at least one memory, and at least one processor configured to generate, based on first analysis on a pathological slide image, first biomarker expression information, generate, based on a user input for updating at least some of results of the first analysis, second biomarker expression information about the pathological slide image, and control a display device to output a report including medical information about at least some regions included in the pathological slide image, based on at least one of the first biomarker expression information or the second biomarker expression information.

Core Innovation

The invention relates to a digital pathology system implemented on a computing device that performs first analysis on a plurality of pathological slide images using a machine learning model. The system generates first biomarker expression information about the pathological slide images based on the first analysis, calculates a biomarker expression score using the first biomarker expression information, and determines priorities of the pathological slide images by comparing the biomarker expression score to predetermined ranges using boundaries for dividing at least three biomarker expression classes.

The system controls a display device to output the pathological slide images based on those priorities. The system further supports updating the analysis results through user input, where a second analysis is performed on a pathological slide image using the user input to update results from the first analysis, and second biomarker expression information is generated based on the second analysis.

The display device outputs at least one of the second analysis results or the second biomarker expression information. In some implementations, the second analysis reanalyzes an entire region of pathological slide images using the user input, and the system calculates a biomarker expression score for cells in a user-drawn region of interest (ROI) based on user input for drawing the ROI.

Claims Coverage

Independent claims are provided for a computing device and for a method. Across the independent claims, the core coverage includes first analysis using a machine learning model to generate biomarker expression information, priority determination based on a biomarker expression score mapped to predetermined class boundaries (at least three classes), and output of pathological slide images based on the priorities. Additional coverage includes user-input-driven second analysis and ROI-based score calculation.

The independent claim set covers machine-learning-based first analysis producing biomarker expression information and a biomarker expression score, priority assignment using predetermined ranges with boundaries dividing at least three biomarker expression classes, and output of pathological slide images based on those priorities. Additional independent claim coverage includes user-input-driven second analysis to generate updated biomarker expression information for display output, and calculating biomarker expression scores for cells within a user-drawn ROI.

Stated Advantages

Documented Applications

No documented applications found