Tempus Launches AI-Powered Paige Predict for Biomarker Detection in Digital Pathology
核心洞察
Tempus AI has launched Paige Predict (搜索), an AI-powered digital pathology solution that predicts biomarkers from H&E slide images when tissue samples are insufficient for molecular profiling.
The platform analyzes 123 biomarkers and oncogenic pathways across 16 cancer types, helping clinicians prioritize testing when tissue availability is limited.
Built using data from over 200,000 patients, Paige Predict (搜索) addresses the critical challenge of "quantity not sufficient" samples that can delay patient care by days or weeks.
Tempus AI has launched Paige Predict (搜索), a cutting-edge artificial intelligence solution designed to predict biomarkers from digital pathology images when tissue samples are too limited for comprehensive molecular testing. The launch follows Tempus's acquisition of Paige (搜索) in 2025 and represents a significant advancement in addressing one of oncology's most persistent challenges: insufficient tissue availability for genomic profiling.
Addressing Critical Tissue Limitations
The growing demand for next-generation sequencing (NGS) and immunohistochemistry (IHC) testing has intensified the challenge of limited tissue availability in cancer diagnostics. When samples are classified as "quantity not sufficient" (QNS), patients face delays of days or weeks while waiting for repeat biopsies or alternative testing approaches. These delays can prolong diagnostic uncertainty and potentially impact treatment decisions.
Paige Predict (搜索) addresses this challenge by analyzing hematoxylin and eosin (H&E) whole slide images to predict the likely presence or absence of clinically actionable biomarkers directly from a single slide. This capability enables physicians to gain insights even when tissue samples are insufficient for full molecular profiling.
Comprehensive Biomarker Analysis
The AI-powered platform leverages Tempus and Paige (搜索)'s intelligent digital pathology technology to identify critical biomarker information from scarce tissue amounts. Paige Predict (搜索) analyzes H&E images to predict the likelihood of 123 biomarkers and oncogenic molecular pathways across 16 cancer types, including non-small cell lung cancer (NSCLC (搜索)), prostate, breast, pancreatic, and colorectal cancers.
The system's ability to predict biomarker presence allows clinicians to inform the sequence in which they order confirmatory tissue-based tests, maximizing the likelihood of obtaining actionable results before exhausting available tissue. Results are automatically delivered with clinical reports to ordering physicians.
Robust Development and Validation
Paige Predict (搜索) was developed using Paige (搜索)'s foundation model and a combined, multimodal cohort containing de-identified data from over 200,000 patients from both Tempus and Paige databases. The model underwent rigorous validation to demonstrate performance, generalizability, and robustness across multiple diverse datasets, including a large-scale cohort from Tempus.
"Tissue can be scarce, but insights don't have to be," said Ezra Cohen, MD, Chief Medical Officer, Oncology at Tempus. "Paige Predict (搜索) overcomes one of the biggest barriers in NGS by delivering actionable information to inform critical testing decisions for patient care when sequencing isn't feasible. This is a defining moment as Tempus and Paige (搜索) continue our work together following last year's acquisition, demonstrating the transformative technology our teams are bringing to digital pathology."
Precision Medicine Integration
Tempus positions itself as a technology company advancing precision medicine through artificial intelligence applications in healthcare. The company maintains one of the world's largest libraries of multimodal data and operates a system designed to make that data accessible and useful for clinical applications. The integration of Paige Predict (搜索) represents the company's continued effort to provide AI-enabled precision medicine solutions that facilitate personalized patient care while supporting therapeutic discovery and development.