{
  "id": 7146945,
  "title": "AI automates 3D membrane mapping, matching manual results in a fraction of time",
  "url": "https://urgent.news/2026/09/13/ai-automates-3d-membrane-mapping-matching-manual-results-in-a",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-13T18:00:06.000Z",
  "source": {
    "name": "Phys.org",
    "slug": "phys-org",
    "url": "https://phys.org/news/2026-09-ai-automates-3d-membrane-manual.html"
  },
  "original_language": "en",
  "account": "Membrane mapping in cells has long been a slow and manual process that relies on 3D images, making it challenging to study cellular processes at the molecular level. Researchers from Helmholtz Munich, the Technical University of Munich, and the Biozentrum of the University of Basel have developed MemBrain v2, an AI tool that automates this task, reducing the time required from weeks to mere hours.\n\nThe software finds membranes, locates specific membrane proteins, and analyzes their spatial arrangement, providing insights into cellular organization. With little to no additional training data, MemBrain v2 can be applied to various research questions without extensive adjustments. This makes it accessible to researchers worldwide, enabling them to study cells faster and on a larger scale.\n\nOne of the main challenges in cryo-electron tomography (cryo-ET), a technique that allows researchers to visualize cells in three dimensions at high resolution, is the presence of gaps in information due to technical limitations. MemBrain v2 addresses this issue by automating membrane segmentation, protein localization, and spatial analysis, thereby overcoming the need for manual annotations or training data.\n\nIn one test, the tool successfully localized protein complexes on additional membranes based on only a single manually annotated membrane, achieving an impressive F1 score of 91%. Previously, this 3D image data had to be painstakingly labeled by hand, making the results difficult to reuse for new datasets.\n\nMemBrain v2 combines three steps into a single AI tool: membrane segmentation (MemBrain-seg), protein localization (MemBrain-pick), and spatial analysis (MemBrain-stats). This comprehensive approach has proven to be as accurate as manual analyses while being significantly faster. The tool's open-source components have already found widespread use, including in research from the Chan Zuckerberg Imaging Institute.\n\nBy accelerating and democratizing membrane analysis, MemBrain v2 enables researchers to study cellular processes across larger datasets, potentially leading to a deeper understanding of how cells function and what changes during disease development. Already, the tool has contributed to new biological insights, such as revealing the spatial separation of photosynthesis proteins within membranes, challenging previous models of their organization.",
  "summary": "Cell membranes and the proteins within them control many vital processes and play a key role in health and disease. But studying them in 3D images of cells has so far meant slow, manual work. A team from Helmholtz Munich, the Technical University of Munich (TUM) and the Biozentrum of the University of Basel has developed MemBrain v2, an AI tool that automates this task—cutting work that once took…",
  "key_points": [
    "AI tool MemBrain v2 automates 3D membrane mapping in cells",
    "Reduces time from weeks to hours, addresses cryo-ET gaps",
    "Achieves 91% F1 score with minimal training data"
  ],
  "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."
}