Dennis Jen

As a data visualization engineer at Bluefin Labs, I build tools for visualizing our massive amounts of TV, ad, and social media data. This was a big departure in domain from my work at the Broad Instiute of MIT and Harvard, where I built tools for visualizing genomic features on genomes and tools for understanding comparative genomics. Working with users to devise innovative and thoughtful visualizations to bring insight to their investigations is a passion of mine I discovered while getting my M.S. at the University of North Carolina at Chapel Hill. During my tenure there, I published a first author paper at IEEE Conference on Visualization. In this work, I collaborated with neuroscientists to help them understand where calcium was located on small neuronal structures.

After graduating, I worked at the Pacific Northwest National Laboratory, where I took on several interesting projects, ranging from visualizing relationships between a massive number of documents to helping analysts understand fish moving across a local dam and was awarded a key contributor award for my part. At this time, tabletop displays weren't as prevalent as they are now, and I was fortunate enough to work with a small team investigating user interaction with such a device coupled with a large projector display.

I came to Boston to work at the Martinos Center for Biomedical imaging to work on neuro-imaging projects. I worked mainly with diffusion tensor imaging, an extension of magnetic resonance imaging (MRI), to develop algorithms for understanding the flow of blood in the brain. I was even able to participate in a few of the studies and run various analyses on data collected from my brain.



When I'm not hunched over a computer, I'm hunched over a pottery wheel.

[See More]


Argo is the Broad Institute's genome annoation and visualization tool that I helped further develop. The tool, built in Java and Swing, was able to provide whole genome gene density overviews and enable biologists to zero in on areas of interest at a gene level or nucleotide level.

The most challenging project I contributed to was in Argo's comparative visualizations. We were charged with building an interactive tool with a variety of levels of detail for comparing multiple genomes. By using a variety of visual encodings, including color, texture, and shape, we were able to highlight areas of genome insertion, deletion, and mutation.



The goal of the POI-Stats (Path-of-Interest Statistics) algorithm is to calculate the highest probability path between two user-defined seed regions from magnetic resonance diffusion tensor data (to discover pathways in brains). The best path is determined by minimizing the energy of the entire path through randomization of the position of the control points of a spline curve drawn through the data and of the position of the endpoints.

While at the Martinos Center for Biomedical Imaging, I ported this algorithm from its matlab implemention to a C++ version that I integrated into the Insight Toolkit (ITK) and 3D Slicer.