Christian Ledig

Timeline

1. By 2015

   A PhD at Imperial College

   After a diploma in applied mathematics at Friedrich-Alexander University Erlangen-Nuremberg, he earns a PhD in medical image analysis in 2015 in Daniel Rueckert's BioMedIA group at Imperial College London. His thesis on robust multi-structure segmentation of MR brain images quantifies the volume of more than 100 individual brain structures. From 2013 to 2015 he also works part-time as a Senior Imaging Scientist at IXICO, commercialising selected algorithms from his research.

2. 2016

   3D networks for brain lesions

   In 2016 he coauthors, with Konstantinos Kamnitsas, Daniel Rueckert, Ben Glocker and others, a dual-pathway 3D convolutional neural network with a fully connected conditional random field for brain lesion segmentation in MRI. The method reaches top ranking performance on the public BRATS 2015 and ISLES 2015 benchmarks, and Medical Image Analysis accepts the paper.

3. 2016–2017

   Super-resolution at Twitter

   He works at Twitter in London from 2016 to 2017 as a Sr Researcher. His focus is patented image super-resolution techniques that improve the user experience.

4. 2017

   SRGAN at CVPR 2017

   In 2017 he is first author of the CVPR paper that presents SRGAN, a generative adversarial network for single-image super-resolution, with coauthors including Lucas Theis, Ferenc Huszar and Wenzhe Shi. SRGAN trains with a perceptual loss that combines an adversarial loss with a content loss, and it infers photo-realistic natural images at 4x upscaling factors.

5. 2018–2021

   FDA-regulated AI in startups

   From 2018 to 2021 he works in the US startup sector, as Senior Research Scientist at Imagen Technologies in New York and Head of AI at VideaHealth in Boston. He leads research teams that build FDA-regulated, AI-driven medical systems for diagnoses in radiography and dentistry.

6. April 2022

   Chair at the University of Bamberg

   In April 2022 he begins leading the Chair of Explainable Machine Learning at the University of Bamberg's Faculty of Information Systems and Applied Computer Science. He heads xAILab Bamberg, and he is a member of ELLIS and of the management board of the Bamberg Center for Artificial Intelligence (BaCAI).

7. 2024

   The MedMNIST-C benchmark

   In 2024 he coauthors MedMNIST-C with Francesco Di Salvo and Sebastian Doerrich, an open-source benchmark of 12 datasets and 9 imaging modalities. It tests medical image classifiers against realistic corruptions and supplies targeted data augmentation to improve their robustness.

8. As of October 2026

   Leading xAILab Bamberg

   As of October 2026 he heads the Chair of Explainable Machine Learning and xAILab Bamberg, whose research focuses on explainable and robust machine learning, particularly in healthcare. In 2026 the Fränkischer Tag profiles his work on AI for brain diagnostics, cancer detection and digital biopsies of the gastrointestinal tract.