Tumor Purity Calculation Using Noise-Aware Tissue and Cell Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for calculating tumor purity from pathological slide images fail to account for biological and technical noise, leading to inaccurate results due to degraded nucleic acid quality and fragmentation, resulting in high false negative or false positive rates.
Innovation Solution
A method and apparatus that analyze pathological slide images using a computing system to perform classifications on tissues and cells, combining the results to calculate tumor purity while considering noise, thereby providing accurate information on tumor purity and expected cancer signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to calculate tumor purity from pathological slide images, then the calculation process is simple, but the accuracy is low due to failure to account for biological and technical noise
Solution Approach 1:
The patent applies segmentation by dividing the tumor purity calculation into multiple classification stages: first classifying tissues into tumor and non-tumor regions, then classifying cells within those regions into tumor cells and non-tumor cells. This multi-stage segmentation approach enables accurate noise differentiation while maintaining systematic complexity management.
Solution Approach 2:
The patent introduces an intermediary noise classification system that identifies and separates biological noise (non-tumor cells within tumor regions) and technical noise (artifacts and degraded regions) from actual tumor signals. This intermediary classification layer mediates between raw image data and final tumor purity calculation, improving accuracy by filtering out confounding factors.
2Reliability
If noise information is included in tumor purity calculation, then diagnostic accuracy improves, but the complexity of analysis increases
Solution Approach 1:
The patent segments noise into distinct categories (biological noise from non-tumor cells and technical noise from artifacts) and processes each category through dedicated classification pathways. This segmentation allows the system to systematically handle noise complexity while improving diagnostic reliability through comprehensive noise characterization.
Solution Approach 2:
The patent changes the analytical parameters by incorporating noise level assessments and classification confidence scores into the tumor purity calculation framework. By adjusting these parameters and weighting different classification results appropriately, the system achieves improved reliability without overwhelming complexity.
3Measurement precision
If multiple classifications are performed on tissues and cells, then tumor purity information becomes more accurate, but processing time increases
Solution Approach 1:
The patent performs preliminary tissue classification before cell classification, establishing regional boundaries and characteristics in advance. This preliminary action allows subsequent cell-level classifications to be more efficient and targeted, reducing overall processing time while maintaining high measurement precision through the multi-stage approach.
Solution Approach 2:
The patent applies partial action by focusing classification efforts on regions and cells most relevant to tumor purity assessment. Rather than uniformly processing every pixel and cell with equal detail, the system applies varying levels of analysis intensity based on regional importance and noise characteristics, optimizing the balance between accuracy and processing time.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Provided is a computing apparatus including: at least one memory; and at least one processor, wherein the at least one processor is configured to: perform a first classification on a plurality of tissues expressed in a pathological slide image by analyzing the pathological slide image, perform a second classification on a plurality of cells expressed in a pathological slide image by analyzing the pathological slide image, and calculate tumor purity including information on noise included in the pathological slide image by combining a first classification result and a second classification result.