Pediatric Head Models in Development
EGI scientist Sergei Turovets is working with Linda Larson-Prior and Fred Prior of the Washington University in St. Louis School of Medicine to develop much needed pediatric head models. The new head models will cover ages 0 to 18. Drs Larson-Prior and Prior have extensive experience with pediatric neuroimaging. EGI brings dense array EEG (Geodesic EEG System 300), electrode impedance tomography (EIT), and 3D sensor position measurement (Geodesic Photogrammetry System) to the project. Together the team will build head conductivity models for infants, children, and adolescents.
EGI customers continue to demonstrate the superiority of dense array EEG (dEEG) compared to conventional EEG. This comparison is illustrated in EGI’s recent article ‘2009 Clinical Year in Review’ in which we looked at a number of landmark publications on recent epilepsy studies. In many cases, clinicians worked to identify the epileptogenic region by mapping the scalp EEG data onto the cortical surface of the brain. This process of solving the inverse problem requires a comprehensive understanding of the forward solution.
The forward solution, or volume conduction model, maps current flow from its source to the scalp. Developing a realistic head model for an adult is challenging enough. As techniques to map conductivity levels improve, we revise our models. New conductivity measurements have led to a skull/brain ratio of 1/15 rather than the previously accepted ratio of 1/80. Do these refinements help us build a better pediatric head model? Not necessarily. While a better understanding of the adult head model is useful, there are enough differences between adult model parameters and pediatric model parameters, and for that matter with the pediatric age spectrum, to warrant development of head models specific to particular age ranges from 0 to 18.
Some of the differences in adult versus pediatric model parameters are quite evident while others remain subtle and not well understood. In the development of pediatric head models we will look at the following:
The developing skull undergoes rapid growth from birth to two years of age followed by slower growth until around six years. Skull composition also changes during this period, changing from mostly cortical bone at birth to diploic bone sandwiched between layers of cortical bone. While the closure of fontanelles generally occurs within the first two years, ossification of cranial sutures continues into early adulthood.
During the development period from 0 to 18 years, the brain undergoes remarkable changes. Significant changes are noted in overall brain volume, in regional tissue composition, in ventricular volume, and in white matter volume. Perhaps the most challenging area will be in modeling how the changes in fractional anisotropy and mean diffusivity affect the accuracy of source localization.
As the white matter volume increases with axonal myelination, we see changes in fractional anisotropy and mean diffusivity. It is not yet well understood how this will affect the accuracy of source localization.
We anticipate that the Turovets, Larson-Prior, and Prior project will develop five new models within the next few years. These head models will distinguish the developmental stages of 0-2 years, 2-4 years, 4-6 years, 6-10 years, and 10-18 years.