December 1 - 6, 2024
Boston, Massachusetts
Symposium Supporters
2024 MRS Fall Meeting & Exhibit
SB02.09.16

Flexible Neural Probe with a Bioresorbable U-Beam Insertion Shuttle for Deep Brain Decoding in Non-Human Primates

When and Where

Dec 4, 2024
8:00pm - 10:00pm
Hynes, Level 1, Hall A

Presenter(s)

Co-Author(s)

Janghwan Jekal1,Saehyuck Oh1,Kyung-In Jang1

Daegu Gyeongbuk Institute of Science and Technology1

Abstract

Janghwan Jekal1,Saehyuck Oh1,Kyung-In Jang1

Daegu Gyeongbuk Institute of Science and Technology1
Flexible neural probes, which closely mimic the mechanical properties of brain tissue, significantly mitigate immune responses and reduce astrocyte and microglia activity compared to rigid neural probes during long-term implantation. This compatibility enhances the stability of chronic neural recordings, which is critical for brain decoding. Despite these benefits related to immune response, the flexibility of these probes complicates their insertion into the brain regions of non-human primates (NHPs). This is because the force necessary to penetrate the white matter and properly position the probe in the brains of NHPs significantly exceeds that required for rodent models. This discrepancy has resulted in limited research on the use of flexible neural probes in primates, compared to the extensive methodologies developed for rodents. Despite this, non-human primates (NHPs) remain crucial for pre-clinical research. Their complex neural connectivity and resemblance to human brain anatomy make them invaluable for advancing our understanding of neurological disorders and testing new treatments. In this study, we introduce a bioresorbable U-beam shaped insertion shuttle-assisted flexible neural probe with porous electrodes, designed for AI-enhanced deep brain decoding in NHPs.<br/> <br/>The shuttle, crafted from a disaccharide-based material, is engineered with a U-beam shape to minimize brain tissue invasion. It offers a transiently high moment of inertia and a buckling force of 90 mN. This design enables the insertion of neural probes up to several centimeters in length into the primate brain without risk of buckling. Moreover, the U-beam configuration optimizes the moment of inertia while minimizing contact area, resulting in minimal brain tissue damage after 4 weeks of implantation.<br/> <br/>Upon insertion, the shuttle rapidly resorbs to expose the underlying dual-layer neural probe, which features 32 porous electrodes. These electrodes are electroplated with platinum(Pt) / iridium oxide(IrOx) and incorporate a 3D nano-porous structure using nanobeads, achieving a low impedance of approximately 40 kΩ. This configuration enables a high signal-to-noise ratio (SNR) neural signal capture.<br/> <br/>To validate the probe's efficacy, it was implanted in the lateral hypothalamic area (LHA) of NHPs, an area associated with food consumption. We integrated a custom wireless neural recording system into the probe for in vivo testing. Local field potentials (LFPs) were recorded from the LHA during phases with and without food exposure over a period of four weeks. The LFPs were processed using a 4th order Butterworth band-pass filter to eliminate noise, followed by conversion into scalograms for visual analysis. The scalogram data revealed increased gamma activity during food presence. Utilizing this data, a convolutional neural network (CNN) classifier with three convolution layers and two dense layers, optimized with the Adam algorithm, was employed to decode the neural patterns, achieving a classification accuracy of 90.3% in distinguishing the feeding phases.<br/> <br/>In conclusion, the deployment of a minimally invasive, biodegradable shuttle has addressed the limitations of flexible neural probes for deep brain recording in NHPs. The probe enabled effective, month-long wireless recording of LFPs, which were subsequently decoded using a CNN-based classifier to determine eating behaviors. These results underscore the potential of such technologies in advancing brain circuit research and electroceutical development.

Symposium Organizers

Reza Montazami, Iowa State Univ
Jonathan Rivnay, Northwestern University
Stephen Sarles, Univ of Tennessee-Knoxville
Sihong Wang, University of Chicago

Session Chairs

Reza Montazami
Stephen Sarles

In this Session