Particle.news

Download on the App Store

Mount Sinai Team Debuts MARQO, an AI Tool for Rapid Whole-Slide Cancer Analysis

The research-use system processes intact stained slides on standard GPUs within minutes.

Image

Overview

  • The peer-reviewed study in Nature Biomedical Engineering details MARQO from a team led by Sacha Gnjatic at the Icahn School of Medicine at Mount Sinai.
  • The pipeline analyzes entire tumor slides without patching and completes runs in minutes on standard graphics cards, avoiding costly compute clusters.
  • MARQO supports common immunohistochemistry and immunofluorescence staining protocols to improve comparability and reproducibility across studies.
  • It automatically flags likely positive cells and records coordinates and marker intensities, with final determinations made by pathologists.
  • The tool is not validated for clinical diagnostics, and the team plans UI refinements, advanced spatial and neighborhood analytics, and high‑performance computing deployments for large slide cohorts.