Medany, M. | Search bioRxiv

bioRxiv2025

Ultrasound-induced Particle Dynamics in Pathological Vascular Vortices

Disturbed flow is a hallmark of diseased vasculature, yet its influence on particle behavior under external actuation remains poorly understood. We uncover distinct behaviors of microparticles under disturbed flow when exposed to ultrasound, revealing selective trapping and aggregation phenomena that differ fundamentally between soft and rigid particles. Using microfluidic models of disturbed vascular flow, we show that microbubbles become trapped at the eye of vortices and self-assemble via ultrasound-induced forces. As clusters grow to a critical size, they are ejected and adhere to the wall opposite the ultrasound source, forming nuclei that progressively occupy aneurysm cavities--a mechanism that could enable targeted, noninvasive treatment. These findings reveal an unexplored interplay between ultrasound and hydrodynamic forces, offering a new strategy for ultrasound-guided therapeutic delivery in vascular disease. One-Sentence SummaryVortices are common in diseased arteries, yet we dont know how therapeutic carriers behave in them under ultrasound. We discover that microbubbles self-assemble in vortex cores, then eject and anchor to vessel walls--revealing a new transport regime with therapeutic potential.

bioengineering↗

bioRxiv2024

Model-Based Reinforcement Learning for Ultrasound-Driven Autonomous Microrobots

AI has catalyzed transformative advancements across multiple sectors, from medical diagnostics to autonomous vehicles, enhancing precision and efficiency. As it ventures into microrobotics, AI offer innovative solutions to the formidable challenge of controlling and manipulating microrobots, which typically operate within imprecise, remotely actuated systems--a task often too complex for human operators. We implement state-of-the-art model-based reinforcement learning for autonomous control of an ultrasound-driven microrobot learning from recurrent imagined environments. Our non-invasive, AI-controlled microrobot offers precise propulsion, which efficiently learns from images in data-scarce environments. Transitioning from a pre-trained simulation environment, we achieve sample-efficient collision avoidance and channel navigation, reaching a 90% success rate in target navigation across various channels within an hour of fine-tuning. Moreover, our model initially successfully generalized in 50% of tasks in new environments, improving to over 90% with 30 minutes of further training. Furthermore, we have showcased real-time manipulation of microrobots within complex vasculatures and across stationary and physiological flows, underscoring AIs potential to revolutionize microrobotics in biomedical applications, potentially transforming medical procedures.

bioengineering↗