Bahetihazi Maidu
Welcome to my personal webpage!
I am a PhD student at the Mechanical Engineering Department at the University of Washington under the supervision of Dr. Juan Carlos del Alamo. I obtained my master's degree at the Mechanical and Aerospace Engineering Department at the University of California San Diego where I specialized in computational modeling and cardiovascular fluid dynamics.
My PhD research centers on the development of biomedical image–based, physics-informed deep learning approaches for reconstructing flow fields and stratifying cardiovascular disease risk from incomplete medical imaging data, with a specific focus on stroke risk assessments in the left atrium and left ventricle.
Outside of work, I like hiking, taking road trips, and enjoying nature with families and friends.
My Recent Research and Publication Snippet
Color-Doppler ultrasound imaging measures 1D flow velocity towards and away from the transducer on a 2D plane, missing detailed flow information and pattern. Intraventricular vector flow mapping (VFM) technique has gained traction in the past decade to reconstruct 2D flow maps using color-Doppler images by integrating mass conservation equation. However, this technique does not consider momentum balance (e.g., Navier-Stokes equations) and remain sensitive to imaging artifacts.
In our most recent work, we developed AI-VFM leveraging recent advances in physics-informed deep learning to reconstruct left ventricular 2D velocity and pressure maps. 3D extension in color-Doppler and applications of this work in computed tomography (CT) imaging are in progress.
