PowerML

VISTA-3D

Sandbox

Interactive 3D medical image segmentation — automatic and point-prompt workflows for CT and MRI.

image segmentationinteractive annotationBayer internalmedical imaging
Hosted on PowerML (AWS eu-central-1)

Pick an example scan or upload a NIfTI file. Scrub slices with the slider and overlay segmentation masks after running inference on the GPU backend.

Input

NVIDIA / MONAI · 132 CT classes — automatic + interactive point prompts.

AutomaticChoose organ class below
Point labelsClick image to add +/- points (VISTA-3D only)
Class pickerPick organ — VISTA-3D labels
ModalitiesCT BODY

Active example: Abdomen CT

Add Segmentations

VISTA-3D uses MONAI label indices (e.g. bone=21, lung=20, liver=1).

No segmentations yet. Run Segment to add overlays (multiple classes/models stack).

Output
Active scan: Abdomen CT

Point prompts use Slider view. Switch to Automatic for mouse browse.

Loading Abdomen CT example from server...
Slice:50
Window/Level:
No image
AXIAL
No image
CORONAL
No image
SAGITTAL
Loading scan dimensions...Zoom: 100%