한빛사논문
Young-Tae Kwon1,6, Yun-Soung Kim1,6, Shinjae Kwon1, Musa Mahmood1, Hyo-Ryoung Lim1, Si-Woo Park2, Sung-Oong Kang2, Jeongmoon J. Choi3, Robert Herbert1, Young C. Jang3,4, Yong-Ho Choa2 & Woon-Hong Yeo1,4,5,*
1George W. Woodruff School of Mechanical Engineering, Institute for Electronics and Nanotechnology, Georgia Institute of Technology, Atlanta, GA 30332, USA.
2Department of Materials Science and Chemical Engineering, Hanyang University, Ansan 15588, South Korea.
3School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA 30332, USA.
4Wallace H. Coulter Department of Biomedical Engineering, Parker H. Petit Institute for Bioengineering and Biosciences, Georgia Institute of Technology, Atlanta, GA 30332, USA.
5Neural Engineering Center, Flexible and Wearable Electronics Advanced Research, Institute for Materials, Institute for Robotics and Intelligent Machines, Georgia Institute of Technology, Atlanta, GA 30332, USA.
6These authors contributed equally: Young-Tae Kwon, Yun-Soung Kim.
*Corresponding author
Abstract
Recent advances in nanomaterials and nano-microfabrication have enabled the development of flexible wearable electronics. However, existing manufacturing methods still rely on a multi-step, error-prone complex process that requires a costly cleanroom facility. Here, we report a new class of additive nanomanufacturing of functional materials that enables a wireless, multilayered, seamlessly interconnected, and flexible hybrid electronic system. All-printed electronics, incorporating machine learning, offers multi-class and versatile human-machine interfaces. One of the key technological advancements is the use of a functionalized conductive graphene with enhanced biocompatibility, anti-oxidation, and solderability, which allows a wireless flexible circuit. The high-aspect ratio graphene offers gel-free, high-fidelity recording of muscle activities. The performance of the printed electronics is demonstrated by using real-time control of external systems via electromyograms. Anatomical study with deep learning-embedded electrophysiology mapping allows for an optimal selection of three channels to capture all finger motions with an accuracy of about 99% for seven classes.
논문정보
관련 링크
연구자 키워드
관련분야 연구자보기
소속기관 논문보기
관련분야 논문보기
해당논문 저자보기