The Effect of Parameter Estimation Methods in Modeling Glioblastoma Growth Prediction

Glioblastomas are the most common and among the deadliest brain tumors. Accurate multiphysics modeling of the tumor is crucial and could aid the experts in choosing the best therapeutic strategy. This study aims to examine the different approaches and requirements for modeling the growth of the tumor and demonstrate the differences in short-term and long-term modeling. To validate the accuracy of our results, we used data from two patients diagnosed with GBM. The performance and accuracy of our models were evaluated by comparing the predicted results with actual MRI scans from these patients. Our primary results showed that although all models could illustrate tumor growth in different parts of the brain, the best results were obtained using the patients' MRI series, which resulted in the lowest modeling errors for both patients. Patients who underwent full resection were better modeled, with an error rate lower than 10 percent. For short-term comparisons, using T2 MRIs was more accurate than using T1Gd MRIs.

The Effect of Parameter Estimation Methods in Modeling Glioblastoma Growth Prediction Keywords:

Brain Tumors , Glioblastoma , Modeling , Patient-Specific , Survival Time