Juan Toscano Institute: Brown University Date: September 5, 2024 Title: Inferring in vivo murine cerebrospinal fluid flow using artificial intelligence velocimetry with moving boundaries and uncertainty quantification Abstract: The cerebrospinal fluid (CSF) flow is crucial for clearing metabolic waste from the brain, a process whose dysregulation is linked to neurodegenerative diseases like Alzheimer’s. Traditional approaches like particle tracking velocimetry (PTV) are limited by their reliance on single-plane 2D measurements, which fail to fully capture the complex dynamics of CSF flow. To overcome these limitations, we employ Artificial Intelligence Velocimetry (AIV) to reconstruct three-dimensional velocities, infer pressure and wall shear stress, and quantify flow rates. Given the experimental nature of the data and inherent variability in real systems, robust uncertainty quantification (UQ) is essential. Towards this end, we have modified the baseline AIV architecture to address aleatoric uncertainty caused by noisy experimental data, enhancing our measurement refinement capabilities. Additionally, we implement UQ for epistemic uncertainties arising from both the physical models and the network representation, which includes testing multiple governing laws and exploring various representation models and initializations. Our approach termed Artificial Intelligence Velocimetry with Uncertainty Quantification (AIV-UQ), not only advances the accuracy of CSF flow quantification but also adapts to other applications that use Physics-Informed Neural Networks (PINNs) to reconstruct fields from experimental data, providing a versatile tool for fluid dynamics and beyond.
