Regulatory Science for Engineering Intuitive, Engaging, Safe and Effective Human-Device Interaction
(Sanjay Yengul, FDA)
Background: Transcranial Ultrasound Stimulation (TUS) is a growing neurosurgical and neuromodulation approach in which computational modeling is often used for treatment planning. Several modeling tools are available for performing TUS simulations, including k-Wave in MATLAB and BabelBrainViscoFDTD in python. Accurate and reliable simulations are needed to ensure both safety and effectiveness of TUS procedures, and to better understand the bioeffects of ultrasonic stimulation. Model accuracy depends on the validity of the governing equations, the numerical methods employed, the discretization used, and the accuracy of the model's input parameters. The input parameters include the acoustic and thermal properties of tissues such as the scalp, the skull, the cerebrospinal fluid, the white and grey matter in the brain, etc. These properties which vary with age and condition, and may therefore be patient specific, are often not known accurately, and have complex dependencies on frequency, temperature, etc. Efforts are underway to develop a reliable database of tissue parameters. A key step in this direction to understand the relative importance of each of the model's input parameters to the end goals in the therapeutic treatment planning process.
Research Plan: The primary goal of this project is to develop a TUS model using existing state of the art tools and study the sensitivity of the model's outputs to the uncertainty in the model inputs. Model inputs include the acoustic and thermal properties of the tissues, such as the speed of sound, ultrasonic attenuation, the acoustic impedance, the specific heat capacity, etc. Model outputs include the characteristics of the acoustic and thermal fields generated in the targeted brain region, such as the focal location, the focal size, peak and average pressure amplitudes, peak temperature rise, the thermal dose achieved, etc. Uncertainty estimates will be derived for the model outputs based on the uncertainty present in the input parameters. Model convergence studies will be conducted to verify and validate the results of the model before the sensitivity analysis. Results from the sensitivity analysis with be documented and analyzed to identify the most important input parameters needed for TUS modeling.
Prerequisites: Engineering coursework in (1) Math, (2) Physics, (3) numerical modeling or scientific computing, and (4) 1 or more years of programming experience in MATLAB and/or Python. Knowledge of solid mechanics, acoustics, or heat transfer is a plus.
Dr. Sanjay Yengul is an acoustician and an engineer-scientist with expertise in ultrasound physics, mathematical modeling, and acoustic measurements. He does regulatory research at the U.S. FDA as a Sr. Staff Fellow in the Medical Acoustics group at the Center for Devices and Radiological Health (CDRH). Before the FDA, he worked in industrial R&D for about two decades, in the areas of acoustics and structural dynamics. His Ph.D. research at Boston University and Brigham and Women's Hospital was in medical ultrasound, specializing in shear wave elastography and inverse problems. His bachelor's and master's degrees are both in mechanical engineering, from IIT Bombay and U. of Michigan, Ann Arbor, respectively.