Back in 2017 Katie Bouman, PhD student at MIT gave a TED talk on “How To Take a a Picture of a Black Hole.” Yesterday, Katie’s research project came to fruition when the first generated image of a black hole was created.
The now 29-year-old graduate shared her breakthrough on Facebook yesterday afternoon: "Watching in disbelief as the first image I ever made of a black hole was in the process of being reconstructed.” In 2017 Katie received her PhD in Electrical Engineering and Computer Science from MIT following a B.S.E in Electrical Engineering from the University of Michigan, Ann Arbor, MI in 2011 and an S.M. degree in Electrical Engineering and Computer Science from MIT in 2013. Whilst studying for her PhD, Katie led the team that created the algorithm that helped capture this first-of-its-kind image. The black hole that has been photographed is at the centre of a galaxy called M87.
Black holes are one of the few mysteries left to humankind, and no single telescope is yet powerful enough to capture them. These regions have a gravitational field so intense that no matter or radiation can escape, and are incredibly far away and extremely compact. To help overcome these challenges preventing images being captured, Katie and her team worked tirelessly to assemble data gathered by the Event Horizon Telescope, eight radio telescopes positioned around the world. This data, which was actually gathered two years ago, was taken over a four day period in optimum weather conditions. After the data was collected, scientists from across the globe came together to study these extraordinary objects that warp space and time.
The Event Horizon Telescope shared that the black hole ‘Messier’ resides ‘55 million light-years from Earth and has a mass 6.5 billion times that of the Sun.’ The EHT observations use a technique called very-long-baseline interferometry (VLBI) which synchronises telescope facilities around the world and exploits the rotation of our planet to form one huge, Earth-size telescope observing at a wavelength of 1.3 mm. VLBI allows the EHT to achieve an angular resolution of 20 micro-arcseconds — enough to read a newspaper in New York from a sidewalk café in Paris.’
Katie’s algorithm was combined with several others from leading experts in image processing, data science, AI and many more techniques to finally create three scripted code pipelines to generate the image. The data collected from the eight telescopes was noisy and sparse and needed to be verified over and over again using machine learning to train the algorithms to ensure that the final output was accurate. The multiple algorithms were trained to ensure that eventually, they all generated the same output image. Katie explained that due to the rigorous training and combination of algorithms, the result that Albert Einstein predicted over 100 years ago was the real deal: "No matter what we did, you would have to bend over backwards crazy to get something that wasn't this ring." Leading the imaging section of the project, Katie explained that ‘"No one of us could've done it alone. It came together because of lots of different people from many backgrounds."