Computer Vision News - May 2018

Anne Carpenter is an Institute Scientist and Director of the Imaging Platform at the Broad Institute. Anne, can you tell us about your work? My lab is focused on getting the most out of biological images as possible. This spans from producing software like CellProfiler – which is open-source software used for processing biological images of all sorts to measure particular features that people care about – all the way to using deep learning for profiling images. This means to extract an unbiased fingerprint of how a cell appears, so that we can compare cell populations against each other to see what changes have occurred. I understand that it was a very clear intention of yours to focus on the medical field. Yes, definitely. My training is as a cell biologist. My PhD is in cell biology from the University of Illinois. My lab, however, is entirely computational. We focus on the computational aspects, but we’re completely driven by solving important biological problems. You entered this field because of your training, but was the desire to help solve human problems another influence? Absolutely, my career path has been driven by what will impact the world, and that’s what got me into biology. I had a choice really, as a smart person interested in biology, between going to medical school and helping patients directly, or going into research science. I was excited about the prospect of research science being catalytic. That the impact could be much more than helping a few hundred patients a year, I could impact thousands or even millions by helping to create new treatments and new approaches to doing science that could help others to discover drugs for various disorders. Definitely, the impact on human health is a major driver for me. Is there one major breakthrough that you have observed since you first started to study in this field? I think I’ve witnessed two major breakthroughs. One is a bit more cultural and the other is a bit more technical. The cultural shift that I’ve seen is biologists’ willingness to be more quantitative in how they deal with images. It was not so very long ago, just a decade or two, when biologists would look by eye through a microscope, they would observe some phenomenon, they would describe it in text, they would include a picture in their paper, and that would be the end of the story. Now it’s much more common that biologists have a sense that if they can see something by eye, Anne Carpenter 24 Friday ISBI DAILY Keynote BEST OF