Christian Ledig
Quick Links
Welcome
Welcome to my webpage. Here you can find some information about myself and research projects I am involved in. » Read More
Accepted Position at University of Bamberg
I will join the University of Bamberg as full professor starting April '22. Focus will be on explainable ML and its applications (esp. healthcare). I am thrilled and very grateful for the opportunity to pursue what I am passionate about in my home region.
CVPR 2021: Outstanding Reviewer
Excited to be on the list of outstanding reviewers for CVPR 2021. Thank you for the recognition.
Joined VideaHealth in Boston as Head of AI
In March I have joined VideaHealth in Boston and glad to now have
completed the move to Cambridge.
I am very excited being part of Videa's team and mission to bring AI into dentistry and
ultimately improve the dental health of millions of people.
--> We are hiring!
Generative Image Translation for Data Augmentation
Our work on “Generative Image Translation for Data Augmentation of Bone Lesion Pathology” was accepted at MIDL 2019. Full article available as [pdf]. This is the follow-up of the [abstract] that we presented at Medical Imaging meets NeurIPS.
MALPEM-ADNI: Large biomarker analysis and shared morphometry database for ADNI
We created and analysed segmentations of 138 anatomical regions of 5074 brain MRIs from ADNI. We share all masks, segmentations, features. The full article was published in Scientic Reports: “Structural brain imaging in Alzheimer's disease and mild cognitive impairment: biomarker analysis and shared morphometry database” [doi]
[dataset]
[details]. This work was done a while ago during my time at Imperial in the Biomedia group.
Joined Imagen Technologies / Move to NYC
Very excited to join the team at Imagen Technologies, Inc. in Manhattan, New York to work towards better healthcare by reducing diagnostic errors and improving access to high quality medical diagnoses.
TBI study published & LondonCV Meetup Talk online
Excited that our study on “Regional brain morphometry in patients with traumatic brain injury based on acute- and chronic-phase magnetic resonance imaging” was now published in PLOSONE.
[doi]
[pdf]
My recent talk on “Pushing the envelope of super-resolution” held within the LondonCV Meetup is now available on YouTube. [talk]
Move of webpage
This webpage is now available under christianledig.com.
Also please note my updated email contact. I will not have access to my Imperial email address for much longer...
CVPR 2017
Aloha! It was great catching up at CVPR in Hawaii. A recording of the oral presentation of our work around SRGAN is online here (including slides): [talk]
[slides]
NeuroImage: Clinical
Our paper on “Five-class Differential Diagnostics of Neurodegenerative Diseases using Random Undersampling Boosting” was accepted for publication in NeuroImage: Clinical. Research led by Tong Tong.
[pdf]
New arXiv papers
Very exciting work with Kamnitsas et al. on “Unsupervised domain adaptation in brain lesion segmentation with adversarial networks”
[pdf]
and Caballero et al. on “Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation”
[pdf]
Deep-Medic Paper is out...
Our paper on “Efficient Multi-Scale 3D CNN with fully connected CRF for Accurate Brain Lesion Segmentation” was accepted in Medical Image Analysis.
[pdf]
[github]
Publications update
Our paper on “Group-contrained manifold learning” was accepted in Pattern Recogntion. [doi]
Twitter Cortex Vx, London, UK
In July, I have joined Twitter Cortex as Computer Vision Researcher. We employ and advance machine learning approaches to improve the user's visual experience (Vx). Very recently, we have published an arXiv paper on “Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network” that you can find here: [pdf]
MICCAI and NeuroImage paper accepted
Our work on the “Differential Dementia Diagnosis on Incomplete Data with Latent Trees” was accepted for presentation at MICCAI 2016 and our work on “Instantiated mixed effects modeling of Alzheimer's disease markers” was accepted in NeuroImage. The papers are available here:
[MICCAI - pdf] [NeuroImage - doi]
Publications update
In a recent article we present “A Robust Similarity Measure for Volumetric Image Registration with Outliers” (Snape et al., Image and Vision Computing,
[doi]
[pdf])
Two articles accepted on the modelling and prediction of AD progression
In one article we focus on “Learning biomarker models for progression estimation of Alzheimer's disease” (Schmidt-Richberg et al., PLoSONE, [doi]).
We further propose “A Novel Grading Biomarker for the Prediction of Conversion from Mild Cognitive Impairment to Alzheimer's Disease” (Tong et al., IEEE Transactions on Biomedical Engineering, [doi]).
Enjoy reading.
PredictND: Publication on differential dementia diagnosis
Our article on the “Differential diagnosis of neurodegenerative diseases using structural MRI data” was accepted in NeuroImage: Clinical [doi]
. The work towards this article was done within the predictND project.
Book Release
Our book chapter on the “Semantic Parsing of Brain MR Images” was published:
C. Ledig and D. Rueckert, “Semantic Parsing of Brain MR Images,” In: Zhou, K. ed. Medical Image Recognition, Segmentation and Parsing: Machine Learning and Multiple Object Approaches (1st Edition), Academic Press, pp. 307-336, 2015.
[bib]
[doi]
[pdf]
Imperial student team qualifies for ACM-ICPC World Finals
Congratulations to our student team for their excellent result in this year's NWERC and qualifying for the upcoming ACM-ICPC World Finals in Phuket, Thailand in May 2016.
[news article]
MALPEM segmentation framework released
The whole-brain segmentation framework MALPEM is now available for download. The package performs bias correction, brain extraction and a structural segmentation of a magnetic resonance brain image. A summary PDF report is created. [download] [PDF report]
Challenge: Ischemic Stroke Lesion Segmentation
Congratulations to Kostas Kamnitsas! His deep learning approach was evaluated as a leading method in this year's MICCAI 2015 - Challenge on Ischemic Stroke Lesion Segmentation.
K. Kamnitsas, et al., “Multi-Scale 3D Convolutional Neural Networks for Lesion Segmentation in Brain MRI,” MICCAI Workshop - Ischemic Stroke Lesion Segmentation ISLES, pp. 13-16, 2015, (1st Place).
[pdf]
[bib]
[results]
Publications (Update)
W. Bai, et al., “ Multi-atlas segmentation with augmented features for cardiac MR images,” Medical Image Analysis, 19(1), pp. 98-109, 2015. [pdf]
C. Ledig, et al., “Robust whole-brain segmentation: application to traumatic brain injury,” Medical Image Analysis, 2015, in press. [doi]
European Programming Competition
We just returned from Linkoeping, Sweden from this year's NWERC. Our three student teams did quite well and brought back a Silver medal to Imperial. [news article]
UK/IE Programming Competition
Congratulations to our student teams for their excellent performance. [news article]
MICCAI 2014 (Update)
Our work on the “Manifold alignment and transfer learning for classification of Alzheimer's disease” was selected as Best Paper in the Machine Learning in Medical Imaging workshop.
Our entry in the Computer-Aided Diagnosis of Dementia (CADDementia) challenge ranked 3rd. More information can be found [here].
MICCAI 2014
I am looking forward to attend MICCAI 2014 from September 14-19 in Boston, MA.
Download PEIS
The implementation of the “Patch-based Evaluation of Image Segmentation (PEIS)” paper can now be downloaded [here].
CVPR Video Spotlight
You can find the video spotlight for our poster presentation at CVPR here: [mpeg4][mov].
Webpage created.
Webpage online. No further news yet.