Computing technologies for image operations

Inventors

West, David R. • Stavish, Coleman C. • Yeo, Max • Jackson, Brian H. • Hang, William

Assignees

Proscia Inc

Abstract

A method comprises: receiving, via a processor, an image depicting a tissue; quantifying, via the processor, the image based on: segmenting, via the processor, the image into a plurality of segments; identifying, via the processor, a plurality of histological elements in the segments; forming, via the processor, a network graph comprising a plurality of nodes, wherein the histological elements correspond to the nodes; measuring, via the processor, a feature of the network graph; performing, via the processor, a transformation on the image based on the feature; determining, via the processor, a non-parametric feature of the image based on the transformation; saving, via the processor, the non-parametric feature onto a database.

Core Innovation

The invention describes a networked digital pathology computing system that receives an image depicting tissue, including whole-slide imaging (WSI) tissue images. The system converts high-resolution WSI into a tiled image pyramid using Deep Zoom Image (DZI) and an aggregated-tile archive using ADZI so the images can be accessed efficiently over a network as tiled image tiles requested by clients.

The system performs distributed image analysis using a task manager and worker architecture to extract image features. It quantifies the image by segmenting the image into segments, identifying histological elements in the segments, including nuclei and morphometric features, and measuring non-parameterized transform-based image metrics using transformations such as wavelet transformation, Fast Fourier, Chebyshev transformation, Radon features, Gabor features, and Zernike features.

In parallel, it constructs graph-based tissue architecture metrics by forming a network graph comprising nodes corresponding to histological elements and measuring a feature of the network graph. The image-derived data and external contextual/feedback data are organized into a unified case identifier record stored in a database, and the system delivers the tiled images and unified case identifier data to clients via streaming tile requests. Optionally, tamper-evident, category-specific blockchains are maintained for records.

Claims Coverage

The document includes one independent claim, which defines a pipeline that receives a tissue image, segments it, identifies histological elements, forms and measures a network graph, transforms the image based on the graph feature, derives a non-parametric feature, and saves the non-parametric feature to a database. Dependent claims further specify database implementation, histological element types, transformation types, the nature of the non-parametric feature, and tile-related client data access.

Across the independent claim, the inventive aspects center on linking image segmentation and histological element identification to a network graph whose measured feature drives a transformation from which a non-parametric feature is derived and stored to a database.

Stated Advantages

Documented Applications

No documented applications found