Jul 13, 2022 2:17:00 PM | TCAP Segmentation & Scoring

Learn how TriMetis Life Sciences leveraged machine learning and Ai to recognize and count specific cell types and morphologies through TriMetis Computer Assisted Pathology (TCAP).

Augmenting what we know about FFPE Tissues 


Informed by Pathologists

TriMetis Computer Assisted Pathology (TCAP) takes on the activities that humans are not well-suited for, but are critical to downstream pathology services, molecular testing, and research decisions. 

Activities such as quantifying Spatial Relationships and Counting Pathological Features are more easily analyzed by artificial intelligence.

Board Certified Pathologists have trained the TCAP applications to do the dirty work, enabling Pathologists to focus on the most important activities supporting patient care.

Easily deployed in the pathology workflow, TCAP increases throughput, informs decisions on which cases are ready to test, and others that need review.

Quality Control Ai includes:

  • Normal from Tumor
  • Nuclei count
  • Tumor Concentration Heatmap

Segmentation & Scoring

Cell Segmentation

Going beyond estimates, we leveraged ML/AI to recognize and count specific cell types and morphologies, especially those that take a lot of time to analyze manually. Neural Networks can rapidly identify cells in a fraction of the time it takes a human. We let the computer do the work, so the pathologist can review and approve the results.

Nuclei Count

Percent % tumor is an estimate. With computer assistance, we now can segment and count tumor nuclei. Knowing these results can help inform the pathologist, and affect downstream testing such as Immunohistochemistry and DNA/RNA profiling.

Heatmap Analysis

Knowing where the tissue has the highest tumor concentration helps determine if the entire slide is viable and should be scraped, or if macro dissection of a specific area on the slide.

Future Applications

TriMetis is developing Segmentation and Scoring Applications to assist pathologists with information for decision support. Sign-up to be notified of new Applications.

Leif Honda

Written By: Leif Honda