Artificial Intelligence (AI) is revolutionizing biomedical imaging, but its adoption raises serious questions around transparency, data privacy, and algorithmic fairness. This talk presents a community-driven, open-source extension for the KNIME Analytics Platform that integrates YOLO and Meta’s Segment Anything for high-precision biomedical image segmentation.
Built entirely with Free and Open Source Software (FOSS), this project empowers researchers and clinicians to process sensitive medical images in a transparent and reproducible way—without relying on opaque black-box systems or commercial cloud services. All processing can be executed locally, ensuring compliance with data privacy regulations and promoting digital sovereignty.
The talk will demonstrate how the extension aligns with ethical AI principles:
Transparency: Full access to models, parameters, and processing pipelines.
Fairness: Customizability to adapt models to diverse demographic and clinical datasets.
Security & Privacy: Local execution safeguards patient data, eliminating third-party exposure.
By embedding cutting-edge AI into KNIME—a widely used open-source platform—this project fosters a collaborative ecosystem where researchers can share workflows, reproduce results, and collectively advance ethical biomedical AI.
The presentation will conclude with a live demo and an invitation for the FOSS community to contribute to this ongoing effort.
