Research

Papers

  • Schalk, Worrell, Mivalt, Belsten, Kim, Morris, Hermes, Klassen, Staff, Messina, Kaufmann, Rickert, Brunner, Worrell, Miller “Towards a Fully Implantable Ecosystem for Adaptive Neuromodulation in Humans: Preliminary Experience with the CorTec BrainInterchange Device in a Canine Model” in review
  • Belsten, Kopsaftopoulos “Data-Driven Flight State Identification via Time-Series-Informed Features and Convolutional Neural Network” 2021 AIAA AVIATION Forum
  • Belsten, Adamek, Brunner “Hardware Abstraction to Facilitate the Dissemination and Validation of Electrophysiological Experiments” one-page research paper, EMBC 2020

Washington University in St. Louis School of Medicine

Invasive Cortical Stimulation

In the Department of Neurosurgery at WashU, I developed a system to explore cortical areas with electrical stimulation via ECoG and SEEG, which significantly improves the efficiency of collecting CCEP (cortico-cortical evoked potentials) data. This system also provides the technical basis to perform closed-loop cortical stimulation over hundreds of electrode sites. It is currently in use at Mayo Clinic and will be adopted by additional institutions, such as the Stanford School of Medicine.

This system is built on BCI2000, g.tec cortical stimulator, and g.tec switching unit.

National Center for Adaptive Neurotechnologies

At WashU and NCAN I worked on maintaining and improving BCI2000, a general purpose software for brain-computer interfacing. I added support for a variety of bio-signal acquisition and peripheral devices. These devices include:

Bio-Inspired Fly-by-Feel Aerospace Vehicles

Designed a multi-branch one-dimensional convolutional neural network to predict the flight state of fly-by-feel aircraft using data recorded from the surface of an airfoil. Interested in why a specific architecture has better, or worse, results and what those results can tell us about the system’s dynamics and information present at each sensor.

Oat Milk Research

Work in progress… Come back soon.