I never would have expected that I would be directly contributing to the development of new healthcare technologies while dabbling a bit into the social sciences - yet here I am.
Data Science and Statistics
Medical imaging is a staple in modern medicine, yet their annual error rates range from 3-5%. Gian investigated how automated computing methods, like machine learning, can be used to improve the accuracy and confidence of such technologies. What Gian assessed was the ability of neural networks from Python’s sci-kit learn library to identify testing data with varying amounts of applied noise. This research lays the groundwork for developing future technologies that can provide healthcare workers with a diagnostic aid to reduce medical imaging error, and ultimately, to save lives.