Flexible, scalable, and generalizable photogrammetric head modeling for high-density diffuse optical tomography

Project Summary: EsperImage, a startup company out of Washington University (WUSTL), will develop efficient and flexible data registration and modeling procedures to complement their high fidelity, wearable, optical technology that transcends limitations of both previous optical neuroimaging and functional magnetic resonance imaging (fMRI) tools to provide naturalistic brain imaging in adults and children. A neurodevelopmental disorder affecting 1/36 children in the general population, autism spectrum disorder (ASD) presents highly heterogeneous phenotypes; however, core features are impaired development of social-communicative skills plus repetitive behaviors and restricted interests. Additionally, early difficulties with visuo-motor integration and motor imitation may be strongly associated with impairments of social communication widely reported in individuals with ASD. Because early interventions in toddlers with ASD have been proven to result in improved outcomes, innovative methods for early detection of the alterations in brain function underlying ASD prior to manifestation of behavioral symptoms are necessary to advance treatment strategies and improve prognoses. Current brain mapping methods such as fMRI offer promising sensitivity yet pose significant methodological challenges in studies of awake, interacting, and moving children due to the loud, constraining environment and susceptibility to motion-induced artifacts. Optical neuroimaging provides a potential surrogate for fMRI. However, image quality with traditional optical technology, functional near infrared spectroscopy (fNIRS), systems had been lacking in comparison to the gold standard of fMRI. Development of high-density diffuse optical tomography (HD-DOT), a tomographic version of fNIRS, has dramatically improved image quality, and maps sensory, motor, and cognitive networks with fidelity comparable to fMRI throughout the outer 1 cm of the cortex. Despite these advances, application of HD-DOT to naturalistic studies in children has been limited by large opto-electronic consoles and bulky fiber optics. Several wearable fiber-less fNIRS instruments are becoming available commercially, but all have multiple deficits, including a lack of automatable data registration and modeling such as those proposed herein. This Phase-I STTR project will develop commercialization-ready, robust, & efficient data registration and modeling tools for WHD-DOT devices that match performance of fiber-based HD-DOT for use in pediatric studies of neurodevelopmental disorders. Long Term Impact: The data registration and WHD systems together will enable a better understanding of disease mechanisms as well as monitor response to therapy in the developing brain. The Aim 1 goal is to develop flexible 2D-camera-based photogrammetric array-anatomy registration. The Aim 2 goal will be to develop photogrammetric atlas-derived anatomical models. We will validate the registration and modeling methods in healthy adults and autistic/non-autistic children to establish feasibility through mapping distributed brain activity in untethered participants. This will establish an essential data registration and modeling product for WHD-DOT commercialization development in a follow-up Phase-II proposal.