SwiftMR

SwiftMR is a standalone medical imaging software solution that processes MR images in DICOM format to enhance image quality through noise reduction and increased sharpness. It is intended for use by healthcare professionals in imaging centers, clinics, and hospitals to assist clinicians in reviewing and analyzing MR images across all body parts at field strengths from 0.25T to 3.0T.

SwiftMR uses a deep learning algorithm for noise reduction (adjustable levels 0-8) and a sharpening filter (adjustable levels 0-5) to process DICOM images. The device supports images from conventional vendor reconstruction methods and vendor-implemented deep learning reconstruction pipelines, and is compatible with multiple MRI manufacturers including Siemens, GE, Philips, Canon, Esaote, Fonar, and Fujifilm.

Not stated in this summary.

The subject device is substantially equivalent to the predicate device (SwiftMR K230854) because both share the same intended use (MR image acceptance, enhancement, processing, review, analysis, communication, and transfer), the same input/output format (DICOM), the same intended users (healthcare professionals), and equivalent enhancement functions (noise reduction and sharpening). Validation testing demonstrated that predetermined performance acceptance criteria were met across multiple MRI manufacturers, field strengths, anatomical regions, and protocols.