| K-number | K242166 |
| Device name | TribusConnect |
| Applicant | Tribusmed Beheer BV |
| Product code | QIH |
| Device class | Class II |
| Decision date | Mar 31, 2025 |
| Decision | Substantially Equivalent |
| Regulation | 892.2050 |
TribusConnect is a cloud-based medical image management and processing software that allows healthcare professionals to securely access, review, manipulate, measure, and visualize CT images in DICOM format from remote locations. It is designed for pre-procedural planning and post-procedural analysis of cardiac interventions in adult patients (22 years and older), using machine learning-based heart segmentation and left atrial appendage localization to support clinical decision-making.
TribusConnect is a web-based cloud application supporting multiple platforms (Windows, Mac, iOS, Android) that processes CT DICOM data server-side to produce 2D, 3D, 4D, MPR, MIP, and volume renderings with ML-based segmentation of seven heart structures. It provides manual measurement tools (distance, area, angle, signal intensity) accurate within 5% and semi-automated LAA localization. Unlike the primary predicate (RemotEye Viewer), it supports only CT modality; unlike the secondary predicate (TruPlan), it segments more cardiac structures and uses the LAA tool as navigation rather than measurement initiation.
IEC 62304:2015 (medical device software lifecycle), ISO 14971:2019 (risk management), NEMA 3.1–3.20 (2016) DICOM standard, FDA guidance on Device Software Functions, Cybersecurity in Medical Devices, and Technical Performance Assessment of Quantitative Imaging.
TribusConnect is substantially equivalent because it shares the same intended use (web-based remote viewing and measurement of DICOM images for cardiac intervention planning) and core technological functions with both predicates. Performance validation showed heart segmentation achieved 100% correct bone removal and 100% clinically usable rendered views; LAA localization achieved 97% accuracy within 10mm of the predicate device's landing zone. Manual measurement validation via phantom studies met the 5% accuracy criterion. The differences in supported modalities (CT only vs. multi-modality), number of segmented structures (seven vs. four), and LAA tool use (navigation vs. measurement initiation) do not introduce new safety or performance risks, as confirmed by verification and validation testing and cybersecurity assessment.
View the full FDA submission: accessdata.fda.gov