Intended Use

Automatic labeling, visualization, and quantification of segmentable brain structures from non-contrast head CT scans to aid physicians in assessing intracranial hyperdensities, lateral ventricles, and midline shift.

Technology

Software-only device using a locked AI machine learning algorithm to process non-contrast head CT scans, producing DICOM outputs with segmentation overlays and volumetric measurements of intracranial hyperdensities, lateral ventricles, and midline shift. Hosted on a cloud server, outputs are sent to PACS for physician review.

Performance

Clinical testing compared algorithm outputs to radiologist ground truth, showing mean absolute errors below set thresholds (7.5 mL for hyperdensity volume, 3 mL for ventricular volumes, 2 mm for midline shift). Dice scores for segmentation of hyperdensities and ventricles exceeded 70%, demonstrating high agreement. Performance was consistent across clinical sites and imaging parameters.

Predicate Devices

No predicate devices specified

Device Timeline

  • 1

    Submission

    8/7/2023

    6 months
  • 2

    FDA Approval

    2/5/2024

Other devices from Viz.ai, Inc.

Ready to Sharpen Your Edge?

Join hundreds of your peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.