MIT Graduate Student Leads Algorithm Development Behind First Black Hole Image
Computer scientist Katie Bouman leads the creation of the CHIRP algorithm, which successfully processes massive amounts of telescope data to produce the first-ever image of a black hole.
Computer scientist Katie Bouman leads the development of a groundbreaking algorithm that makes the first-ever image of a black hole possible. The algorithm, named CHIRP (Continuous High-resolution Image Reconstruction using Patch priors), combines data from eight radio telescopes across the globe as part of the Event Horizon Telescope collaboration to create a cohesive visual representation.
Because astronomical signals reach the various radio telescopes at slightly different times, Bouman adopts a clever algebraic solution to cancel out atmospheric noise. By multiplying the measurements from three telescopes instead of just two, the algorithm eliminates extra delays and ensures accurate calculations while extracting the necessary visual information.
The project involves researchers from MIT’s Computer Science and Artificial Intelligence Laboratory, the Harvard-Smithsonian Center for Astrophysics, and the MIT Haystack Observatory. The team ships half a ton of hard drives containing the massive amounts of collected data to the MIT Haystack Observatory, where CHIRP reconstructs and refines the data into the final historical image.