We're thrilled to announce a significant enhancement for medical imaging workflows: Multiplanar Reconstruction (MPR) Annotation Support is now live on Datature Nexus! This powerful feature allows you to seamlessly view and annotate volumetric medical data (DICOM, NIfTI) across synchronized Axial, Sagittal, and Coronal planes, plus a 3D view.

Highlights:

• Comprehensive Visualization: View and annotate anatomy simultaneously from multiple standard anatomical planes.
• Efficient Editing: Modify annotations in any view, with changes reflected instantly across all planes, streamlining 3D segmentation.
• Dedicated Tools: Utilize familiar bitmask tools (Paintbrush, Polygon, Magic Fill, etc.) plus the new Voxel Brush for efficient 3D painting across slices.
• Medical Data Support: Native handling for DICOM series and NIfTI files.
• Model Support: Train state-of-the-art 3D segmentation models like SegResNet, Swin UNETR, and nnUNet directly on the platform.

Learn more about MPR

• MPR Documentation
• How to Upload
• Importing Annotations
• Supported Model

Getting Started:

• Create a new Project with the type 'Medical 3D Detection'.
• Upload your compatible DICOM or NIfTI data following the new guidelines.
• Open an asset in the Annotator to explore the MPR interface and tools!

This update significantly enhances Datature's capabilities for complex medical imaging analysis. Explore the new documentation to get started, and let us know your feedback!

We're excited to announce the launch of Bitmask annotation support in Datature Nexus! This powerful new feature enables highly precise, pixel-level segmentation, perfect for complex objects, datasets requiring fine-grained detail, or instances where bounding boxes and polygons fall short.

Highlights:

• Pixel-Level Precision: Annotate with ultimate control, defining exact boundaries pixel by pixel.
• Complex Shapes: Easily handle intricate object outlines, disjointed regions (like separate parts of the same object), and objects with internal holes.
• Dedicated Tools: Utilize specialized tools designed for bitmasking: Paintbrush, Freedraw, Magnetic Lasso, and Magic Fill.

Read the full guide: Bitmask Documentation

To start using Bitmask annotations:

• Create a new Project in Nexus.
• Select "Bitmask" as the annotation type during setup.
• Upload your data and explore the new tools in the Annotator!

We believe Bitmask annotations will unlock new possibilities for high-accuracy computer vision models. Dive into the documentation to learn more, and as always, we welcome your feedback!