PATMA: parser of archival tissue microarray

Area of Science:

• Digital Pathology
• Computational Biology
• Histopathology

Background:

• Tissue microarrays (TMAs) are crucial for cancer tissue evaluation and histopathological analysis.
• Scanning TMAs generates virtual slides that require individual core extraction for detailed analysis.
• Identifying and ordering cores from virtual slides is challenging due to the lack of direct labeling.

Purpose of the Study:

• To develop an automated procedure for identifying and extracting individual tissue cores from scanned TMA images.
• To create software that facilitates further image processing of single cores for scientific research.
• To improve the efficiency of TMA data preparation for computational analysis.

Main Methods:

• A novel procedure was designed for automatic and semi-automatic core extraction from archival TMA images.
• The method relies on the spatial arrangement of cores to distinguish and separate them.
• The software processes whole virtual slide images to isolate individual tissue cores.

Main Results:

• The automated selection correctly extracted 89% of tissue cores.
• Manual correction, when needed, allowed for full preparation of a whole slide image in approximately 2 minutes.
• The technique significantly reduces the skill and time required for processing large TMA datasets.

Conclusions:

• The developed procedure effectively automates the extraction of individual cores from TMA whole slide images.
• This software streamlines the workflow for researchers conducting histopathological analysis on TMAs.
• The method offers an efficient and user-friendly solution for preparing TMA data for advanced computational studies.