{
  "id": 20641372,
  "title": "Researchers compare AI tools for analyzing tumor tissue",
  "url": "https://urgent.news/2026/11/18/researchers-compare-ai-tools-for-analyzing-tumor-tissue",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-18T18:20:08.000Z",
  "source": {
    "name": "Medical Xpress",
    "slug": "medical-xpress",
    "url": "https://medicalxpress.com/news/2026-11-ai-tools-tumor-tissue.html"
  },
  "original_language": "en",
  "account": "Researchers from Skoltech, the University of Sharjah, and Tokai University have published an open-access review comparing computer vision algorithms for analyzing tumor tissue in histopathology. These algorithms can help identify tissue features in cancers such as lung, breast, colorectal, and prostate. The review highlights challenges in digitized tissue sections, where adjacent tumor regions can differ significantly in cellular composition and structure. To address these challenges, computer vision can quantify features like cell and nucleus shape, tumor-immune cell arrangement, and tumor-to-tissue boundaries. The review compared five tools on publicly available tissue images for four cancer types, finding that Cellpose and StarDist had the highest mean Dice scores in four out of six comparisons. The Dice score measures how closely a tool's segmentation matches expert annotations, with 1 indicating complete agreement. Cellpose and StarDist performed well in low-density triple-negative breast cancer tissue, while StarDist sometimes missed nuclei in densely packed colon cancer images. The authors emphasize the importance of examining slide regions that contribute to model outputs and testing clinical utility across multiple institutions, taking into account variations in staining and scanning methods.",
  "summary": "Researchers at Skoltech, the University of Sharjah in the United Arab Emirates and Tokai University in Japan, together with colleagues, have published an open-access review of computer vision in histopathology. The review examines how algorithms can identify tissue features in lung, breast, colorectal and prostate cancers.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}