Predict genotype earlier

Predict genotype earlier

Assess tumor genomics faster and less invasively by predicting genotype directly from a single H&E slide—no extra tissue needed and virtually zero turn-around time (TAT).
Boost trial enrollment

Boost trial enrollment

Use AI-based pre-screening from H&E to enrich target patient population for efficient and faster enrollment.
Discover resistance markers

Discover resistance markers

Generate AI-based genomic signatures from H&E to uncover new biomarkers, response indicators, and resistance mechanisms.

Rapid AI-Powered Prediction of Tumor Genotype from H&E

Lunit SCOPE GP is an AI-powered digital pathology solution that enables rapid prediction of actionable tumor characteristics from a single H&E slide.

It supports the development of customized pre-screening tools, co-mutation screening strategies, and advanced analysis for translational research.

How Lunit SCOPE GP Supports Your Team

ROC curve showing AI model performance for tumor genotype prediction with internal validation AUC of 0.880 and external validation AUC of 0.846.

Develop customized pre-screening tools

Develop custom pre-screening tools tailored to your needs. Choose between high-sensitivity or high-specificity models with improved predictive power and flexibility.
Screen for actionable tumor characteristics from a single slide

Screen for actionable tumor characteristics from a single slide

Use AI-powered analysis of a single H&E slide to identify multiple co‑mutations—such as EGFR and KRAS G12C—to accelerate and streamline clinical research.
Scatter plot showing correlation between fibroblast density and TGF-β genomic signature prediction score with R=0.7, indicating strong association.

Advanced genomic signature prediction analysis for translational research

Generate AI-derived genomic signatures from H&E to support the discovery of novel biomarkers, therapy response indicators, and resistance mechanisms—such as the TGFβ signature.

Lunit SCOPE GP key features

Key features include genotype prediction probability score, visualization of genotype predicted area, and flexibility for rule-out or rule-in test.
Screenshots showing genotype prediction probability scores, mapped prediction areas, and adjustable sensitivity and specificity settings for AI-based genotype analysis.

Leveraging proprietary large-scale Foundation Models to improve predictive power

Our streamlined internal platform, powered by Lunit’s data-efficient foundation model, enables rapid development of new prediction models with reduced data requirements and shorter turnaround times.
Diagram showing integration of foundation model-based feature extraction with biological feature extraction to enhance AI prediction accuracy.

Case Study:
Co-development
with AstraZeneca

Lunit and AstraZeneca are co-developing an AI tool using Lunit SCOPE Genotype Predictor to analyze H&E slides and rapidly assess the likelihood of NSCLC driver mutations such as EGFR, ALK, ROS1, MET, KRAS, HER2, and others. This cost-effective screening can deliver results before molecular testing, helping clinicians prioritize high-probability samples for faster, more efficient care.

Case Study: Co-developmentwith AstraZeneca

AI Pathology Software Built to Support Biopharma Innovation

Our digital image analysis software helps biopharma companies scale oncology R&D, advance biomarker strategies, and optimize trials. Lunit is built for end-to-end CDx development, with mature regulatory capabilities, broad scanner and file type compatibility, and scalable deployment across digital pathology platforms and CRO partners.
A Breakthrough Solution to Drive Success in Patient Identification

A Breakthrough Solution to Drive Success in Patient Identification

Equip your team with genotype prediction software that accelerates biomarker discovery, trial enrollment, and genomic insights.