Hi, I’m Ella.
I am a Biomedical Engineer and entrepreneur who is passionate about reforming the health economy.
Projects and Experience
Building Hydroponic Garden (EGR 101)
Poster presentation explaining the hydroponic garden I helped build and process, including design criteria, prototyping and testing.
Medical Device Prototyping - Lightbox
Technical report explaining how I built a fully functioning lightbox, utilizing CAD, Arduino programming (C++), schematic layout, PCB layout, SPICE modeling and testing.
Modeling Depressive Disorder and Serotonin Syndrome
Co-authored report researching drug-drug interaction of anti-depressant and cough syrup with analytical graphs through Python.
Testing CNN Segmentation of Tumors in Mammograms
As a final project for Duke BME Imaging & Diagnostics course, we trained three U-Net deep learning models to segment breast tumors in mammograms, investigating whether different image preprocessing methods (PCA, CLAHE and median filtering) improves CNN accuracy. We found statistically significant improvement in tumor segmentation (Dice score: 0.614 vs. 0.521) with pre-processing.
Stack: Python · PyTorch · OpenCV · scikit-learn
Building a Rate-Adaptive Pacemaker from Scratch
In BME Medical Instrumentation, we designed and built a fully functional VVIR pacemaker from scratch, spanning analog circuit design, signal processing, and embedded programming. The device senses cardiac activity, calculates real-time R-R intervals via Arduino, and delivers stimulation pulses when a missing heartbeat is detected, dynamically adjusting pacing rate from 60 to 120 bpm, and validated on a live frog electrogram.
Stack: Arduino · Analog Circuit Design · C++
Innovation & Entrepreneurship Portfolio
Digital collection of work, learnings and reflections during my experience as a Duke I&E student.
Business Behind Health at Duke
Co-founder and co-president of Duke’s first student organization focused on educating, uniting and providing pathways for all students interested in health.
Multitaper Spectral Analysis of Sleep EEG in OSA
Conducted independent research through BME Signals & Systems analyzing overnight EEG data from a patient with obstructive sleep apnea By applying different dynamic spectral analysis techniques, I identified peri-REM alpha bursts invisible to clinical technicians. These findings confirmed that when OSA patients do enter REM, their EEG signature matches that of a normal sleeper.
Stack: Python · MNE · SciPy · YASA · AutoReject · Multitaper Spectral Analysis