Jaeb Center for Health Research
Research Specialist — T1Dexi
Nov 2019 – Aug 2021
Python
OpenCV
Pandas
People with type 1 diabetes estimate meal carbohydrate content to accurately dose insulin, yet the protein and fat content of meals also influences postprandial glycemia. We examined the accuracy of macronutrient content estimation via a novel phone app. Participant estimates were compared with expert nutrition analyses performed via the Remote Food Photography Method (RFPM).
Detecting and removing 10,000+ duplicate meal photos using Python's OpenCV package and meal descriptions using Python to reduce manual staffing burden.
Automated file transfers.
Creating a data analysis monitoring tool using Python's Pandas package to ensure timelines are met.
Used linear regression machine learning algorithm in Python for outlier detection.