Utilizing VR we are able to stimulate each eye separately (top); We integrate a fully conformal system onto the subject (bottom)

Video Demo of test subject using stimulus text speller in response to prompt

Project Description: A text-speller human-machine interface for individuals with severe motor disability, enabled by novel split-eye asynchronous stimulation by VR headset.

Results:

• Using a VR headset, we are able to generate complex steady-state stimuli, with separate frequencies for each eye.
• Using such an interface allows for many classes with high throughput (>240 bits/min)

Role:

• Conceptualized, planned and managed all aspects of project: designed custom circuits, firmware, software, data processing, machine learning & completed systems integration
• Devised and studied a new BCI paradigm: split-eye asynchronous stimulus using a VR headset

Skills/Tools used:

• Project Conception, System Design & Integration
• AutoCAD (2D design for microfabrication)
• Electronic design: EEG analog front-end design, Bluetooth LE
• Firmware development: nRF52832 microcontroller
• Serial protocols: SPI, I2C
• Unreal Engine 4.26: Developing stimulus arrays, virtual environments, text speller, and video game interface
• MATLAB, Python: Data analysis, visualization; preprocessing methods
• Tensorflow: Designing CNN-based classification scheme
• Android, Windows: Cross-platform application development for real-time data processing and machine learning
• Replica molding, PDMS, epoxy processing
• Femtosecond micromachining laser
• Academic literature writing

Publication:

• Unpublished, In Review: Biosensors and Bioelectronics

Illustration of flexible stethoscope device worn on body (top), and device layers & construction (bottom)

Capture of high quality stethoscope data patients with various abnormal pulmonary sounds

Project Description: A chest-mounted digital stethoscope patch which allows for real-time classification of abnormal lung sounds

Role:

• Drafted circuit schematic and first draft PCBs for stethoscope device
• Performed data analysis, preprocessing and machine learning models for classification of pulmonary sounds

Skills/Tools used:

• MATLAB: Analysis of time-series stethoscope data, use of wavelet analysis, spectrography
• Tensorflow: Design of CNN models for classification of abnormal pulmonary sounds

(caption)

Video Demo of test subject using motor imagery to play a rhythm-type game

Role:

• sfh

Summary:

• [TODO]
• [TODO]

Skills/Tools used:

• Project Conception, System Design & Integration
• [TODO]

TODO: Overview photo

(caption)

Video Demo of test subject using motor imagery to play a rhythm-type game

Role:

• Conceived project concept - designed custom circuits, firmware, software, data processing, machine learning & completed systems integration

Summary:

• [TODO] Using a VR headset, w
• [TODO] Using such an interface allows for many classes with high throughput (>240 bits/min)

Skills/Tools used:

• Project Conception, System Design & Integration
• [TODO]

TODO: Overview photo

(caption)

Video Demo of test subject using motor imagery to play a rhythm-type game

Role:

• [TODO]

Summary:

• [TODO]
• [TODO]

Skills/Tools used:

• [TODO]
• [TODO]