In many medical imaging applications, data fitting is an essential post-procession or analysis step. Common examples are ADC calculations from a set of diffusion weighted MRI images and pharmacokinetic analysis of dynamic contrast enhanced (DCE) CT or MRI. ModelFit provides an infrastructure within MITK for voxelwise fitting of 4D data sets.
The fitting routine, fitting infrastructure and result representation are separated from concrete models and fitting strategies and thus it can be utilized for any model fitting task on 4D data, regardless of imaging modality, fitting domain (temporal, spectral, etc.) or mathematical model. This abstraction allows for implementation of own models for respective use-cases by the user, whilst not having to deal with the overhead of the fitting routine. The embedding within MITK enables the user to perform fitting analysis within an eco-system of medical image processing in combination with all other relevant processing steps, such as image registration or segmentation.
The following ready-to-use applications in form of MITK Workbench plugins are offered:
- A general purpose fitting tool using formula parsing
- Pharmacokinetic analysis for DCE MRI using compartment models
- Pharmacokinetic analysis for dynamic PET using compartment models
- Semi-quantitative analysis for dynamic images using non-compartmental approaches
- Voxel-wise fit visualization for evaluation of fit quality
- Fitting of Z-spectra in chemical exchange saturation transfer (CEST) MRI
- T1/T2 Mapping from MRI acquisition with varying echo times
More information about MITK-ModelFit can be found here:
Debus C and Floca R, Ingrisch M, Kompan I, Maier-Hein K, Abdollahi A, Nolden M. MITK-ModelFit: generic open-source framework for model fits and their exploration in medical imaging – design, implementation and application on the example of DCE-MRI
- https://doi.org/10.1186/s12859-018-2588-1 (BMC Bioinformatics 2019 20:31)
Please use this reference for citation if you use the software for your own publications.
Source code & technology demonstration
When building MITK on your own, please be aware that the relevant plugins have to be activated before using the ModelFit features. Using CMake, navigate to your existing build directory and set the MITK_BUILD_CONFIGURATION to MITKModelFitRelease, generate the build system and build the MITK superbuild. This will activate the following plugins:
- ModelFit Generic Data Fitting
- ModelFit Inspector
- Concentration Curve Converter
- Perfusion Curve Description Parameters
- DCE MR Perfusion DataFit
- Dynamic PET DataFit
The publications are based on this source code version.
The Workbench demonstration is based on this source code version: Windows binary download
For information of how to use the plugins offered by MITK-ModelFit in the MITK workbench please see the application online help or the short user manual.
MRI DCE 2 compartment exchange model (2CXM) This data set is used to validate our 2CXM model.
License of testdata
Copyright (c) 2018 German Cancer Research Center, Division of Medical Image Computing and Department of Translational Radiation Oncology. All rights reserved.
This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
Information for attribution
Title of work: MITK-ModelFit dce_2CXM_ValidationDataSet Attribute work to name: German Cancer Research Center, Division of Medical Image Computing and Department of Translational Radiation Oncology Attribute work to URL: http://mitk.org/wiki/MITK-ModelFit Attribute work in publications: Please use the following citation for referencing. Debus C and Floca R, Ingrisch M, Kompan I, Maier-Hein K, Abdollahi A, Nolden M, MITK-ModelFit: generic open-source framework for model fits and their exploration in medical imaging – design, implementation and application on the example of DCE-MRI (arXiv:1807.07353)
- This page was last modified on 30 July 2019, at 08:41.
- This page has been accessed 1,701 times.