Researchers have developed CytoFormer, a novel foundation model for classifying cells in histopathology images. Unlike previous methods that relied on manual pathologist annotations, CytoFormer uses molecular data from spatial transcriptomics paired with H&E staining to supervise cell morphology. This approach, trained on 15.4 million cells across 16 organs, achieved high accuracy and demonstrated superior performance when transferred to new datasets and in active-learning scenarios, offering a more efficient and scalable method for cell-level analysis in routine histology. AI
IMPACT Introduces a novel, more efficient method for cell classification in histopathology, potentially accelerating single-cell analysis and research.
RANK_REASON The cluster contains an academic paper detailing a new model and methodology. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
