CV

Education

• Bachelor of Science in Electrical Engineering, Johns Hopkins University, Aug 2021 - June 2025
• Bachelor of Science in Neuroscience, Johns Hopkins University, Aug 2021 - June 2025
• GPA: 3.82

Research experience

• Neuroengineering & Biomedical Instrumentation Laboratory, Nitish V. Thakor
  The Johns Hopkins University, Aug 2021 – present
  • 3D Visual Feedback System for Myoelectric Motor Training:
    • Developed a Python-based 3D visual feedback system using Linear Discriminant Analysis to enhance control accuracy in myoelectric motor training.
    • Conducted 200+ EMG-based Fitts Law clinical trials on healthy and amputee subjects to assess 3D visual feedback’s impact on prosthesis control.
    • Demonstrated improved pattern recognition control performance in naive operators, enhancing adaptability to complex tasks through motor-based training.
  • Implantable System Wavelet Compression:
    • Developing an implantable system using wavelet compression to minimize motor unit transmission size.
    • Engineered motor unit separation algorithms to isolate motor units and filter out signal contamination.
    • Designing and manufacturing an all-in-one micro recorder, simulator, and wirelessly charged implantable hardware system for rat implant studies.
    • Performing 10+ RPNI and VDMT reinnervation surgeries on rats to study the effects of implantable systems on retaining motor unit information.
• Neuroplasticity & Repair Laboratory, Daniel C. Lu
  University of California Los Angeles, Oct 2022 – Jun 2024
  • Led the development of a comprehensive MATLAB pipeline with a multidisciplinary team to detect and analyze EMG properties in spinal cord injury patients before and after neuromodulation.
  • Pioneered a PCA-based model to detect abnormalities in breathing and gait cycles in spinal cord injury patients, effectively filtering out contaminated trials from the analysis.
  • Designed a filtering algorithm utilizing adaptive template subtraction to remove ECG artifacts and baseline noise from raw EMG signals.
  • Innovatively cross-correlated lung airway pressure with EMG rolling standard deviation to pinpoint activation regions in respiratory muscles.
  • Developed an automated system for detecting evoked potentials in lower limb muscles.
• Laboratory of Mechanical System and Vibration, Xiangyang Zhu
  Shanghai Jiao Tong University, Jun 2023 – Sep 2023
  • Developed an offline system using the Support Vector Machine algorithm for ultrasound, mechanomyography, and Near-Infrared Spectroscopy (NIRS) signals.
  • Conducted tests on EMG-NIRS feature fusion, demonstrating an improvement in offline classification accuracy from 81% to 90%.
  • Developed an embedded C framework to facilitate command transfer from a PC to the myoelectric prosthetic arm’s microcontroller.
• Institute for Artificial Intelligence, Qining Wang
  Peking University, Jan 2019 – May 2021
  • Co-developed an adaptive algorithm for real-time Maximum Dorsiflexion timing detection in robotic transtibial prostheses, improving control across varying speeds and ramp conditions with 100% accuracy.
  • Collected gait data on three amputees at Peking University and EMG + EEG data of biathlon athletes at Dalian Paralympic Training Center.
  • Created muscle training guide for 40+ biathlon athletes.

Publications

Patter Seperability Visual Feedback To Improve Pattern Recogntion Decoding Performance

Levay, G. M., Yang, R., Hunt, C. L., Hodgson, M. C., Kaliki, R. R., & Thakor, N. V. (2024). PATTERN SEPARABILITY VISUAL FEEDBACK TO IMPROVE PATTERN RECOGNITION DECODING PERFORMANCE. Myoelectric Controls Symposium . https://doi.org/10.57922/mec.2515

Projection-based visual feedback of classification outputs improves efficacy of prosthesis training for myoelectric pattern recognition

L´evay, G. M., Yang, R., Hunt, C. L., Hodgson, M. C., Kaliki, R. R., Thakor, N. V. (2023). Projection-based visual feedback of classification outputs improves efficacy of prosthesis training for myoelectric pattern recognition. Paper presented at the American Academy of Orthotists and Prosthetists Annual Meeting.

Maximum Dorsiflexion Detection Based on an On-Board Adaptive Algorithm for Transtibial Amputees With Robotic Prostheses

Xu, D., Yang, Y., Yang, R., & Wang, Q. (2020). Maximum dorsiflexion detection based on an on-board adaptive algorithm for transtibial amputees with robotic prostheses. IEEE Transactions on Automation Science and Engineering, 18(2), 437-447.

Skills

• Coding: Matlab, Python, C, C++，LDA, SVM, K-means, CNN
• Electrical Engineering: PCB Design, KiCad, Arduino, Microprocessors, Signal Systems, Microfabrication
• Biology: EMG/EEG Recording and Analyzing, Motor Unit Spike Sorting, Amputee Clinical Trials, Rat Handling, RPNI/VDMT surgeries

Relevant Coursework

• Bioelectricity from Neurons to Semiconductors
• Signals and Systems
• Computational Neuroscience
• Microfabrication Laboratory

Research Interests

• Neural Interfaces and Neuroprosthetics
• EMG/EEG Pattern Recognition and Signal Processing
• Spinal Cord Injury Rehabilitation
• Motor Unit Signal Compression and Reconstruction
• Multimodal Sensory Feedback in Neuroprosthetics

Academic Reference

• Nitish V. Thakor, PhD (Email: nthakor@bme.jhu.edu)
• Ruyi Huang, PhD (Email: ruyihuang@mednet.ucla.edu)
• Jason Trageser, PhD (Email: jtrages3@jhu.edu)