Encrypted login | home

Program Information

QUATTRO: An Open-Source Software Package for Quantitative Imaging Applications in Assessing Treatment Response

no image available
R Bosca

R Bosca1,2*, V Johnson2 , E Jackson3 , (1) The University of Texas MD Anderson Cancer Center, Houston, TX, (2) The University of Texas Graduate School of Biomedical Sciences, Houston, TX, (3) Texas A&M University, College Station, TX, (4) University of Wisconsin, Madison, WI

Presentations

SU-E-QI-18 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

Purpose:
Quantitative imaging biomarkers are becoming increasingly utilized in early phase clinical trials as a means non-invasively monitoring treatment response. However, such techniques require extensive data pipelines (e.g., image quantitation, image registration, ROI definition, statistical analyses) to capture and model longitudinal parameter changes. The aim of this study was to develop a comprehensive open-source research software environment supporting such analyses with an emphasis on longitudinally assessing treatment response.

Methods:
Using the MATLAB programming environment, an intuitive graphical user interface (GUI) and associated advanced programming interface (API) were developed, providing an imaging study environment collectively titled Quantitative Utility for Assessing TreaTment RespOnse or QUATTRO. Standard tools were incorporated for importing/exporting common data formats such as DICOM 3.0 and associated RT contours, Pinnacle contours, and for performing ROI and voxel-based modeling for a number of common quantitative MR imaging strategies (dynamic contrast-enhanced, diffusion weighted, and dynamic susceptibility contrast, in addition to T₁/T₂ relaxometry). Advanced features for preforming longitudinal analyses were also developed, including image registration (provided by directly linked Insight Toolkit executables), ordinal and logistic regression analyses (utilizing a link with R), and a scripting framework.

Results:
QUATTRO has been successfully implemented and employed in a number of longitudinal phantom studies and quantitative imaging applications. Moreover, the software has facilitated the development and associated feasibility studies of a voxel-by-voxel statistical response model capable of generating probabilistic response maps by incorporating multiparametric quantitative imaging data.

Conclusion:
An open-source software package, QUATTRO, has been developed for applications in assessing treatment response incorporating quantitative imaging. An intuitive GUI provides a front-end tool for utilizing built-in QUATTRO capabilities, while the API facilitates rapid prototyping of future applications, e.g., novel multi-modality analyses, and links to the highly developed libraries of ITK and R enable development of advanced image processing and statistical tools, respectively.


Contact Email: