Fujifilm Corporation · Class II · Cleared Apr 30, 2026
| K-number | K254189 |
| Device name | Synapse 3D Base Tools (V7.2) |
| Applicant | Fujifilm Corporation |
| Product code | QIH |
| Device class | Class II |
| Decision date | Apr 30, 2026 |
| Decision | Substantially Equivalent |
| Regulation | 892.2050 |
Synapse 3D Base Tools (V7.2) is medical imaging software that provides trained professionals with tools to read, interpret, report on, and plan treatment using DICOM-compliant images from CT, MR, CR, US, NM, PT, and XA devices. It performs 2D/3D viewing, image reconstruction, segmentation, measurement, annotation, and specialized analysis for organs including brain, heart, liver, and lungs. It is not intended for primary diagnostic interpretation of mammography images.
The device is substantially the same as its predicate (V7.0) in all core functionalities including 2D viewing, DICOM communication, MPR reconstruction, volume rendering, and reporting. V7.2 adds 15 new applications with automatic or semi-automatic organ extraction using deep learning (Fully Convolutional Networks), including brain subregion analysis, cardiac function analysis, liver analysis, and lung analysis. All new segmentation features meet acceptance criteria with DICE coefficients ranging from 0.73 to 0.99.
IEC 62304 (Medical Device Software Life Cycle); ISO 14971:2019 (Risk Management); DICOM Set (PS 3.1–3.20, 2016); ISO 20417 (Manufacturer Information); IEC 81001-5-1 (Health Software Security); IEC 62366-1 (Usability Engineering); IEC 82304-1 (Health Software Safety).
V7.2 is substantially equivalent because it maintains the same classification, regulatory number (21 CFR 892.2050), product code (QIH, LLZ), and class (II) as the predicate. All core imaging and management functions are identical. New segmentation applications use the same underlying deep learning methodology and achieve clinically acceptable performance metrics comparable to reference devices already cleared by FDA. No new safety or efficacy issues are introduced.
View the full FDA submission: accessdata.fda.gov