Journal article
D. Hormuth, J. Weis, E. Rericha, T. Yankeelov
Medical physics, 2014
Medical physics, 2014
Semantic Scholar
DOI
PubMed
Cite
Cite
APA
Click to copy
Hormuth, D., Weis, J., Rericha, E., & Yankeelov, T. (2014). MO-G-BRF-03: BEST IN PHYSICS (JOINT IMAGING-THERAPY) - Using Quantitative MRI to Test the Validity of the Reaction-Diffusion Equation in Describing in Vivo Glioma Growth. Medical Physics.
Chicago/Turabian
Click to copy
Hormuth, D., J. Weis, E. Rericha, and T. Yankeelov. “MO-G-BRF-03: BEST IN PHYSICS (JOINT IMAGING-THERAPY) - Using Quantitative MRI to Test the Validity of the Reaction-Diffusion Equation in Describing in Vivo Glioma Growth.” Medical physics (2014).
MLA
Click to copy
Hormuth, D., et al. “MO-G-BRF-03: BEST IN PHYSICS (JOINT IMAGING-THERAPY) - Using Quantitative MRI to Test the Validity of the Reaction-Diffusion Equation in Describing in Vivo Glioma Growth.” Medical Physics, 2014.
BibTeX Click to copy
@article{d2014a,
title = {MO-G-BRF-03: BEST IN PHYSICS (JOINT IMAGING-THERAPY) - Using Quantitative MRI to Test the Validity of the Reaction-Diffusion Equation in Describing in Vivo Glioma Growth.},
year = {2014},
journal = {Medical physics},
author = {Hormuth, D. and Weis, J. and Rericha, E. and Yankeelov, T.}
}
Abstract
PURPOSE The reaction-diffusion equation has been used extensively to model brain tumor growth. However, no one has used quantitative, in vivo imaging data to invert the reaction-diffusion equation to estimate model parameters and then use those values to predict future growth. In this contribution first we determine, in silico, the measurement strategies required to accurately estimate model parameters and test how accurately these parameters can be used to predict future growth. We then perform the analogous studies in vivo where quantitative diffusion weighted magnetic resonance imaging (DW-MRI) data is used to estimate cell number to constrain and initialize the reaction-diffusion equation.
METHODS Measurement strategies were tested using an in silico tumor, seeded within a rat brain domain and "grown" for 9 days as dictated by the reactiondiffusion equation. Parameters were estimated from data subsets (using days 1-2,1-3,…1-8) and used to predict subsequent growth. Prediction accuracy was assessed at the region of interest (total volume and cell number, and Dice value) and voxel (concordance correlation coefficient, CCC) levels. Guided by the simulation results, DW-MRI data was acquired in rats (n=12) with C6 gliomas and used to evaluate the model's accuracy for predicting in vivo tumor growth.
RESULTS Both the in silico and in vivo experiments returned error below 15% when predicting total cell number and volume 1-4 days into the future. The in silico study had CCC and Dice values greater than 0.93 for these predictions. However, the in vivo study had lower Dice (0.64-0.81) and quite poor CCC (0.11-0.25) values.
CONCLUSION The simulations show that the reaction-diffusion equation can be used to make accurate bulk scale (i.e., total tumor volume and tumor cell number) predictions of in vivo tumor growth. However, the poor voxel level agreement, suggests the reaction-diffusion equation is an incomplete description of in vivo C6 glioma growth. NCI-1U01CA174706, NCI- P30CA068485, NCI- R01CA138599, NCIR25CA136440.
