Society for Mathematical Biology nautilus logo

International Conference on Mathematical Biology and

Annual Meeting of The Society for Mathematical Biology,

July 27-30, 2009

University of British Columbia, Vancouver

.

Program

Poster PS34B
Misha Simon
University of Washington
Title A comparison between volumetric and localized spatial analysis techniques for assessing model parameters
Abstract Gliomas are the most common type of primary malignant brain tumor, and are characterized by their aggressive growth and uniformly fatal prognosis. Grade IV gliomas, referred to as glioblastoma multiforme (GBM), grow and progress to fatality rapidly, despite aggressive therapy. Because of their diffuse and invasive nature, current medical imaging techniques only capture a fraction of the entire region of malignancy, but mathematical modeling utilizing spatial and temporal information from these images can paint a more informative picture of GBM proliferation and invasion dynamics. Models for GBM aggressiveness focusing on the dynamics the diffuse invasion (net dispersal rate D) and proliferation (net rate ρ) of malignant cells have been shown to accurately reflect the disease progression as seen on medical imaging (Harpold). Swanson et al (Harpold) have demonstrated a methodology for estimating model parameters using a combination of volumetrically-assessed growth velocity and aggressiveness (D/ρ) to accurately model and predict GBM growth and invasion in individual patients. Patient-specific rates of invasion, proliferation and radial growth velocity are assessed and calculated from tumor volumes obtained from MRI scans taken at different time points prior to treatment. This current model parameter estimation methodology relies on gross changes in tumor volume for calculating radial velocity, but for patients with tumors abutting anatomical barriers in the brain or diffusing in non-spherical growth patterns, the use of localized spatial analysis can specifically tailor the model parameter estimates, thus improving the model predictions. To perform the localized spatial analysis, two pre-treatment images for a patient are translated and rotated into spatial correspondence and a minimum Euclidean distance metric is used to assess rates of invasion, velocity and aggressiveness ratio (D/ρ). We can also generate a distribution of velocities of advancement of the imageable tumor front and compare the model parameter estimates for the patient-specific rates of radial growth and biological aggressiveness. The volumetric velocity represents a composite average of the observed range and distribution of velocities and D/ρ produced by the localized spatial analysis method, across various MRI imaging techniques. A quantitative comparison between volumetric techniques and spatially-resolved techniques for model parameter estimation will be provided in this presentation. Analysis reveals that patient-specific predictions of the disease progression will benefit from this detailed anatomical analysis.
CoauthorsRussell Rockne, Rita Sodt, Kristin R. Swanson
LocationWoodward Lobby (Wednesday-Thursday)