Scan Conversion

The Scan Conversion algorithms transform ultrasound images from their native acquisition / reconstruction geometry (such as polar or sector format) into a regular Cartesian grid, making them suitable for visualization, analysis, and further processing. The module provides two main scan conversion algorithms: Advanced Scan-conversion and Scan-conversion for Ultrafast Imaging.

Input

Both algorithms are invoked with ultrasound data that include metadata describing the acquisition geometry (FrameGeometry) and additional ultrasound metadata (UltrasoundMetadata). They can be applied offline to a set of images or online to a stream of images.

Output

The output is a set of ultrasound images resampled onto a regular Cartesian grid, suitable for display or further quantitative analysis.

Description

Scan conversion is a crucial step in ultrasound image formation, as many acquisition systems store data in a geometry that matches the probe’s physical characteristics (e.g., polar or sector format). The scan conversion algorithms use the provided metadata to map each pixel from the acquired geometry to the appropriate location in a Cartesian image.

Algorithm Variants

Advanced Scan-conversion

  • Designed for general ultrasound sweeps, using the FrameGeometry and UltrasoundMetadata to determine the scan conversion parameters.

  • Automatically determines whether to perform forward or inverse scan conversion based on the metadata.

  • For forward scan conversion, the output spacing is determined from the scanconverted image descriptor (if available) or estimated from the prescan data.

  • For inverse scan conversion, the prescan image descriptor must be present in the geometry.

  • Suitable for most conventional ultrasound imaging scenarios.

Scan-conversion for Ultrafast Imaging

  • Specifically supports ultrafast ultrasound imaging, including data that has not been coherently compounded (i.e., a single frame containing multiple plane wave or diverging wave emissions, represented as a 3D image).

  • Also supports scan conversion of coherently compounded data (as 2D pre-scanconverted images).

  • Uses the same metadata and method as ScanConversion2, but is tailored for the unique structure of ultrafast imaging data.

  • Note: Inverse scan conversion is not implemented in this variant.