NIMG-66. A METHOD FOR RAPIDLY CREATING HEAD MODELS OF GLIOBLASTOMA PATIENTS FOR STUDYING THE DELIVERY OF TTFIELDS TO THE BRAIN

Computational studies simulating the delivery of TTFields to realistic head models are a standard method for investigating TTFields distribution in the brain. These studies are useful as they allow estimation of field distributions without the need for invasive physical measurements. Creating realistic head models of patients is labor-intensive because it requires accurate segmentation of patient’s MRI data-sets. This limits the usefulness of standard methods for creating head models in studies that require a large number of realistic head models, such as studies investigating connections between field intensity distributions and disease progression. Here we present a method for creating realistic models of Glioblastoma Multiforme (GBM) patients that overcomes this limitation. A prerequisite for our method is the creation of a highly detailed healthy head model which serves as a deformable template from which patient models can be created. When creating patient models, the tumor in the patient’s MRI is first segmented. Next, non-rigid registration algorithms are used to register the healthy regions of the patient head MRI images on to a 3D image representing the deformable template. This yields a non-rigid mapping of the patient’s head in to the template space, as well as the inverse transformation that maps the template in to the patient space. The inverse transformation is applied to the 3D deformable template and an approximation of the patient head in the absence of a tumor is found. Finally, the segmented tumor is planted back into the deformed template to yield a full patient model. Comparison of the models to patient MRIs reveals good representation of the patient anatomy within the model. This method enables rigorous investigation of important clinical questions related to the connection between TTFields distribution with the brain and disease progression.