Oral Session - New Devices for Intervention & Sensing
Physics & Engineering
713A/B
Wednesday, 07 June 2023 08:15 - 10:15
The MRDust: An Implantable Neural Interface Powered via Focused Ultrasound with Data Communication via MR Image Modulation
Biqi Rebekah Zhao1, Yuhan Wen1, Alexander Chou1, Elad Alon1, Rikky Muller1, Chunlei Liu1, and Michael Lustig1
1Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, Berkeley, CA, United States
We proposed and investigated a novel wireless deep brain implant MRDust as a potential platform for burst-mode neural recording that receives power from focused ultrasound and achieve data communication via dynamic MR image modulation in deep brain at multiple sites simultaneously.
Figure 1: a) Conceptual diagram of implant. It comprises a piezo receiver, two electrodes and an IC. b) Power delivery and downlink control: MRgHIFU directs ultrasonic energy to several areas to power multiple motes at once. Downlink data is encoded in ultrasound by amplitude modulation. c) Burst recording: on command, implants record neural potential in burst mode, amplify, digitize and store them. d) Data uplink: implants encode binary data in dynamic MRI bit by bit, reading recordings from multiple motes in parallel. Blue bar shows order and duration of events in an operation cycle
Figure 2: a,b) Coil current in-sync with GRE-EPI and SE-EPI sequence c) Simulation result of relationship between signal contrast, current and TE in GRE-EPI and SE-EPI for a 10-turn 650um coil in 2x2x2mm3 voxel d) Imaging result of signal contrast, for 4-shot GRE-EPI (TE/TR/FA = 12-80ms/500ms/65°, pixel:1.7x1.7mm2, SL:3.6mm, BW:100Hz/pixel ) & SE-EPI (TE/TR = 17-80ms/2s, pixel:1.7x1.7mm2, SL:3.6mm, BW:100Hz/pixel). In GRE-EPI signal contrast competes with T2* decay.