{"slug": "usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans", "title": "USC researchers publish an Alzheimer's model built for missing scans", "summary": "Tamoghna Chattopadhyay and colleagues at the University of Southern California published MEMOIR-VLM on October 1st, a research model that classifies Alzheimer's disease from whatever combination of MRI, diffusion imaging and clinical scores is available, reaching 91.3% balanced accuracy distinguishing cognitively normal people from people with dementia on held-out research data and 68.2% accuracy across three categories including mild cognitive impairment. The paper's own results show cognitive test scores drove most of the diagnostic performance, and the language model did not beat nearest-neighbor retrieval.", "body_md": "# USC researchers publish an Alzheimer's model built for missing scans\n\n**MEMOIR-VLM combines MRI, diffusion imaging and clinical scores, but its strongest diagnostic signal came from cognitive tests.**\n\n        By [Ryan Merket](https://runtimewire.com/author/ryan-merket)\n        · Published \n\nPrimary source: [X](https://x.com/TamoghnaChatto2/status/2107564670592913807)\n\n## Why it matters\n\nMEMOIR-VLM tests whether one model can handle the incomplete scans common in practice. Its own results show the limit of the pitch: cognitive scores drive most diagnostic performance, and the language model did not beat nearest-neighbor retrieval.\n\n[Tamoghna Chattopadhyay (@TamoghnaChatto2)](https://x.com/TamoghnaChatto2) and colleagues at the University of Southern California published MEMOIR-VLM on October 1st, a research model designed to classify Alzheimer's disease using whatever combination of brain scans and clinical data is available. The paper's headline result was 91.3% balanced accuracy distinguishing cognitively normal people from people with dementia on held-out research data. Its three-category accuracy, including mild cognitive impairment, was 68.2%.…", "url": "https://wpnews.pro/news/usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans", "canonical_source": "https://runtimewire.com/article/usc-memoir-vlm-alzheimers-missing-scans", "published_at": "2026-10-11 06:26:25+00:00", "updated_at": "2026-10-11 07:23:57.238440+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-research", "computer-vision", "natural-language-processing"], "entities": ["University of Southern California", "Tamoghna Chattopadhyay", "MEMOIR-VLM"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans", "markdown": "https://wpnews.pro/news/usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans.md", "text": "https://wpnews.pro/news/usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans.txt", "jsonld": "https://wpnews.pro/news/usc-researchers-publish-an-alzheimer-s-model-built-for-missing-scans.jsonld"}}