4D Flow MRI MATLAB Toolbox

Presenting Author(s)

• 

  Julio Sotelo

  Assistant Professor
  Departamento de Informática, Universidad Técnica Federico Santa María
  Santiago, Region Metropolitana, Chile

Primary Author(s)

• 

  Julio Sotelo

  Assistant Professor
  Departamento de Informática, Universidad Técnica Federico Santa María
  Santiago, Region Metropolitana, Chile

Co-Author(s)

• SU

  Sergio Uribe, PhD

  Professor
  Monash University
  Melbourne, Victoria, Australia
• DH

  Daniel Hurtado

  Pontificia Universidad Católica de Chile
  Santiago, Chile
• JM

  Joaquìn Mura, PhD

  Assistant Professor
  Universidad Técnica Federico Santa María, Chile
• HM

  Hernan Mella, PhD

  Assistant Professor
  Pontificia Universidad Católica de Valparaíso, Chile
• DM

  David Marlevi, PhD

  Postdoctoral Fellow
  Karolinska Institutet, Massachusetts, Sweden
• PL

  Pablo Lamata, PhD

  Professor
  King’s College London
  London, England, United Kingdom
• JF

  Joao Filipe F. Fernandes

  PhD
  King's College London
  London, England, United Kingdom

Description: In this work we present a toolbox developed in MATLAB (MathWorks, Natick, MA, USA), which allows both the scientific and clinical community to process 4D Flow MRI data in a fast and friendly way. To preprocess a 4D flow MRI data, a series of tools can be used such as image offset and noise correction (1), semiautomatic segmentation using the 4D Flow MR angiography image (2), aliasing correction (3). The tetrahedral finite element mesh is generated using the MATLAB toolbox iso2mesh (4). Subsequently, the hemodynamic and geometric parameters shown in Table 1 are calculated. For data saving, the toolbox allows storing these data in different formats, MATLAB, EXCEL (Microsoft Corporation), VTI and VTK (Visualization toolkit).

We also integrate advanced algorithms to estimate pressure as vWERP (5) and SAW (6). We have begun to incorporate algorithms such as 4DFlowNet (7), to obtain super-resolution of 4D Flow MRI images. Additionally, we have also integrated a manual multiplanar reformatting module, which allows the extraction of 2D Flow slices from the 4D Flow MRI data, to estimate a series of flow parameters.

The main window with aortic segmentation in three orthogonal views and corresponding velocity values are shown in Figure 1. Also, the module that creates a tetrahedral finite element mesh was shown in Figure 2. This tool has been applied to patient groups with different conditions like bicuspid aortic valve (8), aortic dissection (9), and transposition of the great arteries (10). Future plans include adding intracardiac 4D Flow MRI analysis and AI-based features, such as automatic vessel segmentation, noise reduction, anti-aliasing, and multiplanar reformatting.

Access Instructions: To access the toolbox, you can download it from the following GitHub repository (https://github.com/JulioSoteloParraguez/4D-Flow-Matlab-Toolbox). This code was developed by Dr. Julio Sotelo (Professor at the Department of Computer Science at Universidad Técnica Federico Santa María). It is compatible with Windows and macOS and can also be used on Linux. To generate the tetrahedral finite element mesh, the iso2mesh toolbox is required (http://iso2mesh.sourceforge.net/cgi-bin/index.cgi?Home).

The 4D Flow MRI dataset must be loaded into the app as a MATLAB structure. This approach was chosen because different vendors frequently modify the DICOM header with each software update, so using a standardized structure ensures compatibility. The data structure and all the requirements necessary to generate it are described in the repository GitHub repository.

Table 1: Hemodynamics and Geometrical parameters that can be quantified in the toolbox.

Parameter	Representation	Classification
Velocity [m/s]	Volume	Hemodynamics
WSS [N/m2]	Surface	Hemodynamics
WSSA [N/m2]	Surface	Hemodynamics
WSSC [N/m2]	Surface	Hemodynamics
Vorticity [1/s]	Volume	Hemodynamics
Forward Velocity [m/s]	Volume	Hemodynamics
Backward Velocity [m/s]	Volume	Hemodynamics
OSI [-]	Surface	Hemodynamics
Helicity Density [m/s2]	Volume	Hemodynamics
Relative Helicity Density [-]	Volume	Hemodynamics
Viscous Dissipation [1/s2]	Volume	Hemodynamics
Energy Loss [uW]	Volume	Hemodynamics
Kinetic Energy [uJ]	Volume	Hemodynamics
Angle in Axial Direction [°]	Volume	Hemodynamics
Regurgitant Flow [%]	Surface	Hemodynamics
Eccentricity [%]	Surface	Hemodynamics
Diameter [cm]	Surface	Geometrical
Radius [cm]	Surface	Geometrical
Curvature [1/m]	Surface	Geometrical
Ellipticity [-]	Surface	Geometrical
Length Vessel [cm]	Surface	Geometrical
Forward Vortex [1/s]	Surface	Geometrical
Flattering [-]	Surface	Geometrical
Area [cm^2]	Surface	Geometrical
Axial Circulation [cm2/s]	Surface	Hemodynamics

Figure 1. Main window of the toolbox, in which we can see three standard views of a 4D flow MR angiography image (axial, coronal, and sagittal), with a 3D representation of the velocity vectors in a finite element mesh of the aorta. In addition, we can see the different buttons available in the toolbox, such as unwrapping, multiplanar reformating, vWERP, segmentation, among others.
Figure 2. Secondary window that allows the generation of the finite element mesh, domain in which the velocity data is transferred, in order to subsequently quantify different hemodynamic parameters.