Accelerate Biomarker Discovery with Multimodal AI
Free Access Available Now - No Download Required
Designed for scientists without AI expertise to develop AI models from histology and real-world data and explore biomarker hypotheses through multimodal insights.
OI Suite helps BioPharma uncover biomarkers and identify the right patients from limited, heterogeneous data. The Discovery package provides free access to CanvOI, trained on 1.5 million H&E-stained tissue samples, enabling rapid, no-code exploration of biomarker hypotheses and predictive models using public real-world datasets
The OI Suite Key Advantages
Rapid, no code model development
Independently develop AI models in minutes using our CanvOI pathology foundation model. Train binary, multi class, and regression models directly from digitized biopsy images. No code or AI expertise needed.
Biomarker discovery with multimodal insights
Generate and refine biomarker hypotheses from morphological, genomic, and clinical insights. Run differential analyses, compare across subtypes, and identify signals that link tissue patterns to molecular features.
Explainable results and visual interpretation
Visualize the histological regions driving each prediction through clear, interactive heatmaps. Review tissue patterns, validate findings with confidence, and share reproducible outputs across teams.
OI Suite Discovery - Selected Use Cases
Talk to us about the Full OI Suite
Our most commonly asked questions
What is OI Suite?
OI Suite is a SaaS enterprise platform designed to address one of oncology development’s most pressing challenges: the difficulty of discovering biomarkers and identifying the right patients for the right therapies from limited data.
OI Suite full package is powered by our Cross-Modality Intelligence Engine integrating histology records, DNA, and clinical text alongside our published State-Of-The-Art histology foundation model CanvOI. CanvOI (co-developed with Oracle) is now trained on more than 1.5 million digitized pathology slides across 40 organs/tissue types. With this pre-training, OI Suite is designed to perform with low-data cohorts (i.e. as few as ~20 cases with response data per indication) to increase predictive power and decision confidence for cohort enrichment and early protocol decisions.
OI Suite offers an intuitive interface that requires no coding or AI expertise, translating our Cross-Modality Intelligence Engine into practical, accessible tools.
What can I do with the OI Suite?
OI Suite offers an intuitive interface that requires no coding or AI expertise. The platform key capabilities includes:
- No-code predictive modeling: Enables researchers to rapidly and independently build models from routine H&E and multi-modal datasets and derive biologically grounded signals to generate and refine biomarker hypotheses.
- Biomarker discovery augmented with RWD: Expands small cohorts with external data to increase prediction power
- Interpretability tools: Heatmaps, tissue and cell segmentation, and feature-level insights trace model predictions back to morphology
- Built for low-data cohorts: Robust results possible from as few as 20 annotated samples. Increase predictive power and decision confidence for cohort enrichment and early protocol choices.
What is the difference between OI Suite Discovery and Full Package?
OI Suite full package is powered by our Cross-Modality Intelligence Engine integrating histology, gene expressions, mutational profiles, and clinical records. OI Suite provides access to Imagene AI Real-World Data lake. Researchers use OI Suite to independently generate biomarker hypotheses, digitally expand indications, and improve patient identification and stratification, enabling evidence-based trial design and cohort enrichment. All research is conducted on de-identified data under rigorous privacy, security, and institutional oversight.
Email us at OISuite@Imagene-ai.com to learn more.
OI Suite Discovery
is Powered by
Our state-of-the-art Oncology Intelligence foundation model co-developed with Oracle and trained on more than 1.5 million tissue samples across 40+ organs and tissue types.
When benchmarked against publicly available foundation models, CanvOI achieved up to 20% higher performance in low-data settings.