Shutterstock DLS is a Big Hit with Faculty and Students
It was standing room only for the second talk in the 2017-2018 Shutterstock Distinguished Lecture Series on November 17. The people who packed lecture hall in the New Computer Science building at Stony Brook University were there for MIT Professor Nir Shavit's presentation, High Throughput Connectomics: The Making of a Brain Scope.
Opening the lecture with a poem titled, The Man Watching, Dr. Shavit entices the audience to think of science in a new light. “This poem perfectly captures the nature of science and the work of a scientist,” Shavit explained. As the lecture continued, Shavit explained connectomics and its place in both neurobiology and computer science. Connectomics takes the principles of neurobiology and mixes the machine learning and processing aspects of computer science. The end result is the extraction of brain connectivity graphs from electron microscopy images.
Dr. Shavit’s talk explored the possibility and feasibility of designing a high-throughput connectomics-on-demand system that runs on a multicore machine. Shavit and his team of researchers have already begun working on producing a similar machine.
The audience bursted with questions for Dr. Shavit surrounding his current research, how the data is analyzed and stored or not stored, the comparison of brain cells to other bodily cells, etc. While trying his best to answer every eager question, Dr. Shavit closed the lecture with a wise metaphor about Galileo’s invention of the telescope. “When Galileo invented the telescope he didn’t know what he’d find,” Dr. Shavit said. “We’re building a telescope with the hope that when we point it to our brains it will show us something like the night sky,” he concluded.
Dr. Shavit teaches computer science at both Tel-Aviv University and MIT, and teaches electrical engineering at MIT. A distinguished professor and writer, Shavit co-authored the book The Art of Multiprocessor Programming. He also won the 2004 Godel Prize in theoretical computer science for his work on applying tools from algebraic topology to model shared memory computability. He is also the winner of the 2012 Dijkstra Prize for the introduction and first implementation of software transactional memory. His current research and field of interest is computational neurobiology. Dr. Shavit’s research centers on developing new ways of using high performance computing to analyze data in order to uncover the microscopic structure and function of neural tissue.
-Author: Samantha Mercado