The Event Horizon Telescope (EHT) can observe distant black holes through a global network of telescopes. However, EHT measurements are sparse and noisy, requiring computational algorithms to recover a black hole image. We developed a course to help students get comfortable with EHT data and image reconstruction algorithms so that they can contribute to EHT research efforts.

The tutorial consits of a series of Colab notebooks along with a Google Slides presentation. We recommend starting with the presentation, which covers relevant theoretical background and guides you through all the code in the notebooks. Overall, the tutorial covers how to use the eht-imaging library and diffusion-model methods to (1) generate synthetic data, (2) reconstruct a synthetic image, and (3) reconstruct an image of the M87 black hole from real EHT data. These steps are a starting point towards more challenging black-hole imaging topics, including polarization, dynamic imaging, and 3D imaging. Such topics present many useful and challenging research questions in the intersection of computer science and astrophysics.

This tutorial was originally created for the ICCP Summer School held on July 19, 2025.

@misc{feng2025eht,
  title={EHT Black-Hole Imaging},
  author={Feng, Berthy T and Bouman, Katherine L},
  howpublished={\url{https://github.com/berthyf96/eht_imaging_tutorial}},
  note={Tutorial first presented at the 2025 ICCP Summer School in Toronto, Canada},
  year={2025},
}