Roche Plays for Bigger Stakes in Digital Pathology with Expanded AI Integration for Cancer Diagnostics
Roche has announced an expansion of its Digital Pathology Open Environment, integrating over 20 artificial intelligence (AI) algorithms from eight new collaborators into its navify® Digital Pathology platform. This move is aimed at improving cancer diagnosis by using AI-driven image analysis tools that support pathologists in analyzing tissue samples more accurately, particularly for cancer-related diagnostics and research. The expansion aligns with Roche’s strategy to leverage its control over upstream hardware and software, offering a platform that facilitates integration with third-party AI tools.
AI Collaborators and Their Focus Areas
The integration includes AI algorithms developed by eight collaborators, each specializing in different aspects of cancer detection and analysis:
- Deep Bio: Algorithms for prostate cancer detection, grading, and tumor quantification.
- DiaDeep: Tools for breast cancer biomarker quantification.
- Lunit: Tumor proportion score (TPS) analysis for non-small cell lung cancer (NSCLC).
- Mindpeak: Algorithms for breast cancer biomarkers and PD-L1 analysis for lung, gastric, esophageal, bladder, and breast cancers.
- Owkin: Detection of microsatellite stability (MSS) in colorectal cancer.
- Qritive: Screening and grading of prostate cancer, analysis of lymph nodes for metastasis, and colon cancer screening.
- Sonrai Analytics: Detection of microsatellite instability (MSI) status in colorectal cancer.
- Stratipath: Risk profiling for invasive breast cancer.
These algorithms are integrated into Roche’s navify® Digital Pathology platform, allowing pathologists to apply specialized AI-driven tools directly within their workflow. Each collaborator’s algorithms focus on specific types of cancer or biomarkers, addressing the need for precise image analysis and quantification in pathology.
Roche’s Digital Pathology Open Environment
Roche’s Digital Pathology Open Environment operates as an ecosystem where external AI tools can be integrated with the company’s existing digital pathology platform. The navify® Digital Pathology platform acts as the interface between Roche’s slide scanning hardware and third-party AI software, facilitating the analysis of high-quality digital images of tissue samples. The open environment allows Roche to create an infrastructure where external AI developers can contribute new algorithms, enabling a modular approach to pathology.
Katie Maloney from DeciBio compares Roche’s approach to an App Store business model, where Roche’s control over the upstream hardware (slide scanners) and platform (image management software) provides leverage for integrating downstream AI tools. This strategy positions Roche as both a provider of digital pathology solutions and a gatekeeper for AI-driven innovations, allowing third-party developers to create AI tools that plug into Roche’s system.
Topic: AI in Bio
