Welcome to multimodars documentation!

multimodars: A Rust-powered cardiac multi-image modality fusion package.

Warning

Not intended for clinical use.

This package aims to combine different cardiac imaging modalities to combined 3D models. A particular focus is on the fusion of intravascular imaging (either intravascular ultrasound [IVUS] or optical coherence tomography [OCT]) with a 3D geometry created from computed coronary tomography angiography (CCTA).

The main functionalities are organized into two major categories:

  1. Intravascular Alignment: This category addresses the co-registration of intravascular frames through translation and rotation to minimize the distance between adjacent frames. This approach is particularly useful for correcting misalignments introduced by cardiac motion during intravascular image acquisition. It further encompasses the alignment of separate intravascular pullbacks (e.g., acquired under rest and stress conditions) as well as the co-registration of systolic and diastolic frames within the same pullback.

  2. CCTA-Intravascular Fusion: This category covers the fusion of intravascular frames with a three-dimensional geometry reconstructed from coronary computed tomography angiography (CCTA). This is particularly valuable for generating high-resolution three-dimensional models of coronary arteries by combining the high spatial resolution of intravascular imaging with the three-dimensional geometry provided by CCTA. Functionalities in this category include the anatomical labeling of CCTA geometries based on centerlines, the registration of intravascular frames onto a centerline, and the subsequent fusion of the resulting geometry with the CCTA reconstruction.

Intravascular image data are accepted either as CSV files following the output format of AIVUS, an open-source tool for automated intravascular ultrasound analysis, or directly as NumPy arrays, affording greater flexibility for integration with custom pipelines.

Table of Contents

License

This package is covered by the open source MIT License.

Developers

1Department of Cardiology, Inselspital, Bern University Hospital, University of Bern, Switzerland 2Graduate School for Cellular and Biomedical Sciences, University of Bern, Bern, Switzerland

Contributing

We’d welcome your contributions to multimodars. Please read the contributing guidelines on how to contribute to multimodars.

Changelog

Indices and tables