3D Acoustic Simulation and Optimization Algorithms for Transcranial Focused Ultrasound Delivered with Stereotactic Robotics

Overview

Neurodegenerative diseases, such as Alzheimer’s disease, are one of the leading causes of disability and death worldwide. Current pharmaceutical treatments for these diseases are hindered by the lack of efficient drug delivery methods across the blood-brain barrier (BBB). Fortunately, focused ultrasound (FUS), a rapidly emerging, noninvasive clinical device, can open the BBB to increase the bioavailability of therapeutics in the brain for improved medical treatment. Although beneficial in various clinical settings, FUS suffers from attenuation and distortion caused by the heterogeneous human skull, which yields a deviation between the focal point of FUS and the target. Finding an optimal placement of the transducer with respect to each subject and desired target is very challenging. In this paper, I define this problem as constrained optimization and develop a novel iterative search algorithm to optimize single-element FUS transducer placement based on accurate 3D acoustic simulations of transcranial FUS propagation. From digital medical images, I present an automatic, universal framework to reconstruct high-fidelity acoustic and geometric properties for precise representations of skull heterogeneity and accurate 3D FUS simulations. Then, I design and implement a novel, graphics processing unit-accelerated iterative search algorithm to optimize FUS transducer placement. My algorithm outperforms the state-of-the-art, achieving a high accuracy and minimal error of 3.85 ± 1.37 mm and 3.56 ± 2.12 degrees. The novel modeling, simulation, and search algorithm are integrated into a surgical robot to establish an end-to-end framework for patient-specific FUS treatment, such as BBB opening for drug delivery and thermal ablation of cancerous tumors.

Competition history

• JSHS 2023 Category not listed

Resources

• National JSHS abstract book ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• JSHS None projects
• Every JSHS 2023 project