Ultrasound Scan Conversion: Pixel Density Is Not Spatial Resolution

By ALEVA | California, USA

Engineering question: What changes when a sector image gets a denser Cartesian output grid, but the beamformed source data stays the same?

Separate sampling from presentation

For a sector scan, scan conversion maps samples indexed by range and angle onto a Cartesian image. Chang, Yen and Shung's peer-reviewed 2008 study examines interpolation choices for this operation. It is foundational work, not a new 2026 product announcement. Interpolation estimates output values from available samples; increasing raster density alone adds no independent measurements. A smoother display is not evidence of better acoustic resolution. [1]

A reproducible geometry example

Assume adjacent beam directions are separated by 1 degree. At a radius of 60 mm, their arc separation is rΔθ = 60 × π/180 ≈ 1.05 mm. A 0.20 mm output spacing puts approximately 5.2 pixels across that angular interval. It does not demonstrate a 5.2-fold resolution improvement. This is a hypothetical geometric calculation, not an ALEVA measurement; beam spacing is not the point-spread function.

Define the resampling contract

An integration specification should record source geometry, units, output spacing, origin, orientation, interpolation and outside-sector handling. ITK's versioned 5.3.0 ResampleImageFilter documentation explicitly separates output geometry, transform and interpolator settings. Its transform is output-to-input. This is a useful general resampling reference, not validation of a particular ultrasound implementation. [2]

The engineering trade-off is fidelity versus implementation cost: an interpolator must be assessed against source sampling, artifacts, memory traffic and latency. A more elaborate method is not automatically preferable for every acquisition or processor.

Proposed OEM acceptance checks

• Hold beamformed input fixed when comparing output grids and interpolators.

• Assess point targets and profiles in physical units at multiple depths and angles; do not compare screen-pixel widths alone.

• Check geometry, sector boundaries and image texture alongside latency, transfers and sustained operation on the intended host.

• Version the coordinate mapping and interpolation settings with each reproducible test.

These are proposed verification checks, not measured ALEVA performance or clinical validation. For an OEM/ODM decision, specify acceptable physical-image behavior and timing before selecting raster size or processor. Scan conversion should not be confused with re-beamforming or a separately validated super-resolution method.

References

1. Chang JH, Yen JT, Shung KK. High-Speed Digital Scan Converter for High-Frequency Ultrasound Sector Scanners. Ultrasonics. 2008;48(5):444–452. DOI: 10.1016/j.ultras.2008.03.001. PubMed record.

2. Insight Toolkit. ResampleImageFilter, ITK 5.3.0 official API documentation. Version-specific documentation, accessed October 19, 2026; not presented as a new release.