Application of whole slide image markup and annotation for pathologist knowledge capture

Area of Science:

• Digital Pathology
• Computational Histopathology

Background:

• Whole slide imaging (WSI) generates large datasets, posing challenges for data storage and transfer.
• Retaining pathologist knowledge through image markup and annotation is crucial for research and education.
• Variations in slide scanning can lead to inconsistencies in image analysis and annotation placement.

Purpose of the Study:

• To investigate the transferability of image markup and annotation data between different scans of the same glass slide.
• To assess the feasibility of using annotation data files to capture pathologist knowledge without retaining entire WSI files.
• To develop a method for overcoming scan variations in digital pathology.

Main Methods:

• Mathematical principles were applied to address variations in WSI scans.
• Trilateration was employed to precisely link fixed points within images and slides.
• Annotation placement was associated with image data using a metadata file.

Main Results:

• Annotation placement variation was reduced from over 80 μ to less than 4 μ on the x-axis.
• Annotation placement variation was reduced from 17 μ to 6 μ on the y-axis.
• The reduction in variation was statistically significant (P < 0.025).

Conclusions:

• The developed methodology enables the creation of reproducible histopathology image libraries.
• This approach facilitates the sharing of interpretations and knowledge for educational and research purposes.
• It offers a solution for efficient knowledge capture and retention in digital pathology.