Reslice3Dto2D: Introduction of a software tool to reformat 3D volumes into reference 2D slices in cardiovascular magnetic resonance imaging

TL;DR

Reslice3Dto2D is a new software tool that reformats 3D cardiovascular MRI data into 2D slices for better compatibility with existing post-processing tools.

Contribution

The tool introduces a two-step interpolation method to reformat 3D data into reference 2D slices with slice profile selection.

Findings

The tool was validated on an artificial dataset and tested on 119 subjects with various pathologies.

Reformatted data could be imported into three different post-processing software tools.

Image sharpness decreased by 40% with 7 mm slice thickness due to partial volume effects.

Abstract

Cardiovascular magnetic resonance enables the quantification of functional and morphological parameters with an impact on therapeutical decision making. While quantitative assessment is established in 2D, novel 3D techniques lack a standardized approach. Multi-planar-reformatting functionality in available software relies on visual matching location and often lacks necessary functionalities for further post-processing. Therefore, the easy-to-use Reslice3Dto2D software tool was developed as part of another research project to fill this gap and is now introduced with this work.

The Reslice3Dto2D reformats 3D data at the exact location of a reference slice with a two-step-based interpolation in order to reflect in-plane discretization and through-plane slice thickness including a slice profile selection. The tool was successfully validated on an artificial dataset and tested on 119 subjects with different underlying pathologies. The exported reformatted data could be imported into three different post-processing software tools. The quantified image sharpness by the Frequency Domain Image Blur Measure was significantly decreased by around 40% on rectangular slice profiles with 7 mm slice thickness compared to 0 mm due to partial volume effects. Consequently, Reslice3Dto2D enables the quantification of 3D data with conventional post-processing tools as well as the comparison of 3D acquisitions with their established 2D version.

The online version contains supplementary material available at 10.1186/s13104-024-06931-4.

Figures