MIGHTY TRENDING
Bill Bostock

This is how scientists captured the first picture of a supermassive black hole

(Screenshot/BBC News)

The algorithms that played a major role in allowing a supermassive black hole to be photographed for the first time were largely designed three years ago by a graduate student in her 20s.

Katie Bouman, now 29, was studying computer science and artificial intelligence at the Massachusetts Institute of Technology, and she worked at the school's Haystack Observatory.


Scientists published the first image of a black hole. The image captured Event Horizon Telescope observations of the center of the galaxy M87.

(Event Horizon Telescope Collaboration)

In the search for a way to capture an image of the black hole, located 55 million light-years away in the heart of the Messier 87 galaxy, astronomers at MIT took part in the Event Horizon Telescope project, but they faced a serious problem.

They needed to stitch together millions of gigabytes' worth of data captured by telescopes located all over the world.

Bouman had the solution: Find a way to stitch the data about the black hole together pixel by pixel.

Katie Bouman.

(TED/YouTube)

"We developed ways to generate synthetic data and used different algorithms and tested blindly to see if we can recover an image,"Bouman told CNN.

"We didn't want to just develop one algorithm. We wanted to develop many different algorithms that all have different assumptions built into them."

"If all of them recover the same general structure, then that builds your confidence."

Vincent Fish, a scientist at MIT's Haystack Observatory, told CNN that Bouman was "a major part of one of the imaging subteams."

Fish told CNN that senior scientists worked on the project too, but the specific task of imaging the black hole was predominantly run by junior researchers like Bouman.

"One of the insights Katie brought to our imaging group is that there are natural images," Fish said.

"Just think about the photos you take with your camera phone, they have certain properties." He added: "If you know what one pixel is, you have a good guess as to what the pixel is next to it."

CNN reported that Bouman would begin teaching as an assistant professor at the California Institute of Technology in the fall.

This article originally appeared on Business Insider. Follow @BusinessInsider on Twitter.