Tumor Necrosis Morphology

Original H&E stained whole tissue section (left) and classified regions of interest (ROI) overlay (tumor – orange; necrosis – blue; blood – green; white space – white) 

Analysis Approach

We develop an algorithm to identify different morphologies characterised by their color, texture and contextual features. Based on these learning characteristics, distinct regions of interest (ROI) can be segmented. The completed algorithm is applied to all tissue sections in the study in an automated and objective manner. Example analysis and data is shown.

     Original H&E stained tissue section (left) and ROI classification overlay (tumor – orange; necrosis – blue; blood – green; white space – white) (right)

Quantitative Readouts

  • Area (mm^2) across a whole tissue section or within a specific ROI.


Example Tumor Necrosis Morphology Data

Benefits

  • The user is able to define their desired ROI for subsequent quantitative IHC analysis.
  • Eliminate the need to manually outline regions to be analysed (edit this).
  • Automated and objective application of one algorithm across a study set will reduce variability and enhance data quality and subsequent interpretation.