Real-Time Whole Slide Cell Counting With Watershed Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current whole slide imaging systems struggle to efficiently count large numbers of cells, particularly one million or more cells, within a selected area of a pathology image, which is necessary for modern cancer diagnostics like NantOmics® Genomic Proteomic and Spectrometry cancer diagnostics, and existing local computer-based methods are too slow.
Innovation Solution
A real-time cell counting system using whole slide stain separation, local drain cell segmentation, and watershed transform methods, integrated within a client-server architecture, to determine cell counts in a selected region of a whole slide pathology image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If local computer-based cell counting programs are used, then cell counting can be performed, but the counting speed is too slow for one million or more cells
Solution Approach 1:
The patent divides the large-scale cell counting problem into multiple smaller processing units by implementing a distributed computing system with multiple worker nodes. Each worker node processes a subset of cells independently, allowing parallel computation across the entire cell population. This segmentation enables the system to count one million or more cells in real-time by distributing the computational load across multiple processors working simultaneously.
2Quantity of substance
If conventional cell counting methods are used for patches with several hundred or thousand cells, then counting can be completed, but the method cannot handle one million or more cells required by modern cancer diagnostics
Solution Approach 1:
The patent transitions from traditional two-dimensional image patch analysis to a multi-dimensional distributed computing architecture. By adding the dimension of parallel processing across multiple worker nodes and implementing hierarchical processing levels (patch-level, slide-level, and aggregate-level counting), the system can simultaneously handle both small patches and large-scale whole slide imaging with one million or more cells, achieving both high quantity and high productivity.
3Quantity of substance
If whole slide imaging is used to capture several million cells, then comprehensive diagnostic data is obtained, but the inability to handle high-throughput cell counting limits widespread adoption
Solution Approach 1:
The patent implements a self-service distributed computing system where worker nodes automatically discover available processing tasks and execute cell counting operations without requiring manual intervention. The system autonomously manages task distribution, result aggregation, and quality control across multiple nodes, enabling the entire pipeline from whole slide imaging to diagnostic-ready cell counts to operate automatically. This self-service capability makes the system easy to use for diagnostic purposes while handling several million cells.
Data Source
AI summary
Techniques are provided for determining a cell count within a whole slide pathology image. The image is segmented using a global threshold value to define a tissue area. A plurality of patches comprising the tissue area are selected. Stain intensity vectors are determined within the plurality of patches to generate a stain intensity image. The stain intensity image is iteratively segmented to generate a cell mask using a local threshold value that is and gradually reduced after each iteration. A chamfer distance transform is applied to the cell mask to generate a distance map. Cell seeds are determined on the distance map. Cell segments are determined using a watershed transformation, and a whole cell count is calculated for the plurality of patches based on the cell segments. A client device may be configured for real-time cell counting based on the whole cell count.