Heather D. Couture, Ph.D.
I have more than 15 years of experience with data-driven image analysis in both academic and industry settings. I have worked on a variety of interdisciplinary R&D projects including digital pathology and planetary science.
I completed a B.Math in Computer Science at the University of Waterloo and, through one of five internships, discovered computer vision. This led me to the Robotics Institute at Carnegie Mellon University for a Masters where I worked on methods to detect and geologically classify rocks from images, an autonomous science technology for planetary rovers and for processing returned images.
Local area startup Digitalsmiths (since acquired by TiVo) drew me to the Research Triangle of North Carolina. While at Digitalsmiths I worked on numerous projects to index movies and TV shows - methods that operate behind the scenes to help consumers find the content they're looking for.
I then returned to academia and completed a Ph.D. in Computer Science at the University of North Carolina at Chapel Hill. During my dissertation work I developed a low cost and repeatable method for predicting breast tumor biomarkers from H&E histology - some of these biomarkers were previously not known to be predictable from images alone. I also explored models for integrating image and genomic data. More broadly, my research in image recognition and machine learning studied methods for forming discriminative representations for large, heterogeneous images and multimodal data. These techniques are applicable to other disease and image types for applications ranging from improving treatment decisions to better understanding the disease itself. They are also extendable to aid discriminability for many forms of heterogeneous data.
I started Pixel Scientia Labs in 2012 to service local companies in need of specialized machine learning and image recognition skills. I have continued to follow the emerging research with deep learning opening up new possibilities in automated analysis of images. Due to my drive to create impact, my business has evolved into a service for R&D teams to maximize their benefit from machine learning.