ROI Annotation Detection in Pathology Slides for Digital Mask Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual annotation of cancerous regions on glass slides by pathologists is time-consuming and expensive, and these pen annotations are often considered artifacts and excluded from computational modeling, limiting the availability of annotated data for AI in pathology.
Innovation Solution
A method to detect and segment hand-drawn pen annotations on digitized pathology slides, converting them into digital masks using an image processing pipeline that includes color space conversion, thresholding, and morphological operations, enabling the extraction of annotation masks for machine learning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual annotation of cancerous regions is performed by pathologists, then annotation accuracy is improved, but time consumption and cost increase
Solution Approach 1:
The system enables automatic self-annotation by detecting pen markings and extracting annotated regions without requiring pathologists to manually create digital annotations. The method processes digitized pathology slides to automatically identify and extract annotated regions, allowing the system to serve itself rather than requiring continuous human intervention for annotation creation.
Solution Approach 2:
The system creates digital copies of pen annotations by detecting the pen markings on digitized slides and generating corresponding digital masks. This copying process transforms the physical pen annotations into digital format, preserving the annotation information while eliminating the need for time-consuming manual re-annotation by pathologists.
2Quantity of substance
If pen annotations are used on glass slides, then annotation availability is improved, but the annotations are considered artifacts and excluded from computational modeling
Solution Approach 1:
The system replaces the mechanical pen-marking process with a computational detection and extraction process. By using image processing techniques to detect pen markings and generate digital masks, the system transforms the mechanical annotation method into a computationally processable format, making the annotations suitable for machine learning and computational modeling.
Solution Approach 2:
The system changes the parameters of the annotation representation by converting pen markings into digital mask images with specific color spaces and pixel values. This parameter transformation converts the annotations from a format considered as artifacts into a standardized digital format that can be reliably processed by computational models and machine learning algorithms.
3Reliability
If manual digital annotation is performed, then computational modeling readiness is improved, but time consumption increases
Solution Approach 1:
The system performs preliminary detection and extraction of pen annotations during the digitization process, before the data is needed for computational modeling. By automatically detecting and extracting annotated regions at this early stage, the system prepares the data in advance, eliminating the need for time-consuming manual digital annotation later in the process.
Data Source
AI summary
The present disclosure is directed to systems and methods that may receive an image, wherein the image includes an annotation at least partially enclosing a region of interest (“ROI”), wherein the image has a plurality of pixels. The systems and methods may use a first algorithm to determine at least one foreground and at least one background from the image. The systems and methods may use a second algorithm to determine a plurality of annotation pixels from the plurality of pixels of the image. The systems and methods may intersect outputs from the first algorithm and the second algorithm to determine an intersection which defines the ROI.