News


Multi-atlas for neonatal rabbit brain

posted Jun 8, 2018, 10:43 AM by Juan Eugenio Iglesias

 

 

We have a new paper in NeuroImage (first author: Sebastian Ferraris) describing a neonatal rabbit brain (multi-)atlas consisting of 12 multi-modal samples, with very precise manual annotations. Rabbit models are becoming increasingly popular in neurodevelopment studies, as they represent a convenient middle ground between small and large animals. The segmentations are are available at the following repository: https://github.com/gift-surg/SPOT-A-NeonatalRabbit
Congrats, Sebastiano, for the great paper!


Two papers accepted at MICCAI

posted May 31, 2018, 5:08 AM by Juan Eugenio Iglesias   [ updated May 31, 2018, 5:31 AM ]

I am delighted to announce that we have got two papers accepted at MICCAI this year (you can find them under Publications). 
  

 
   
In "A probabilistic model combining deep learning and multi-atlas segmentation for semi-automated labeling of histology", my PhD student Alessia Atzeni presents a probabilistic model for semi-automated segmentation of stacks of histological sections, which combines two of the most successful successful families of techniques in medical image segmentation: multi-atlas segmentation and convolutional neural networks. The model has the potential to greatly speed up the manual delineation of stacks of histological sections to create gold standard segmentations. Congrats, Alessia!

 
 


In "Model-based refinement of nonlinear registrations in 3D histology reconstruction", I present a method for recovering the 3D structure of a stack of histological sections, that generates smooth reconstructions while avoiding the common banana effect and z-shift artefacts. The method is based on Bayesian inference within a probabilistic model in the space of 2D nonlinear spatial deformations, parameterised by stationary velocity fields. I have made the code publicly available under SOFTWARE.

Click on the image on the left for an animation illustrating the output of the method!




Atrophy of hippocampal subfields in genetic frontotemporal dementia

posted Apr 26, 2018, 12:24 PM by Juan Eugenio Iglesias

 
 

Our paper "Hippocampal subfield volumetry: differential pattern of atrophy in different forms of genetic frontotemporal dementia" (first author: Martina Bocchetta; senior author: Dr Jonathan Rohrer) has been accepted for publication in the Journal of Alzheimer's Disease. Using our Hippocampal subfield tool, we showed that the hippocampus was affected in subjects with three different mutation types, but also different patterns of subfield involvement depending on the mutation. Moreover, these involvement patters are consistent with cortical-subcortical network vulnerability. 



OHBM Abstract - and award!

posted Mar 9, 2018, 1:04 AM by Juan Eugenio Iglesias   [ updated Mar 13, 2018, 9:20 AM ]

 

Our abstract "Rare genetic events in sporadic Alzheimer’s disease: a network propagation approach" (first author, Marzia Scelsi, senior author, Andre Altmann) has been accepted as an oral presentation at OBHM. 

The abstract presents a method to investigate the effects of rare variants in sporadic Alzheimer's disease cohorts, such as the ADNI, and is based on signal diffusion on a gene-interaction network. Using 800 subjects from ADNI, the results show a robust correlation between the volume of hippocampal subfield CA3 and the smoothed mutation profile of two genes (CLU, OTUD4). In addition, the study also found significant enrichment for these genes and their 179 interactors was found for: genes up-regulated in brains of patients with AD and with incipient AD; neuron development; dendrites; and protein phosphorylation. 

Congrats, Marzia, Andre, et al.!

Update: the submission has won a Merit Abstract Award. Big congrats, Marzia!

Fruit fly brain atlas

posted Mar 5, 2018, 7:16 AM by Juan Eugenio Iglesias   [ updated Mar 5, 2018, 7:22 AM ]

 



My collaborator Ignacio Arganda just published a paper with his former colleagues in France, describing the construction of an atlas of the fruit fly brain (Drosophila Melanogaster); I lent him a hand with the registration and segmentation part.

Over 3000 images from an image collection were registered and are freely available to explore at:
http://www.fruitfly.tefor.net/

The manuscript describing the methods can be found under Publications


Survey paper on 3D histology reconstruction

posted Feb 15, 2018, 12:27 AM by Juan Eugenio Iglesias   [ updated Feb 15, 2018, 12:28 AM ]

 
 


We have just gotten a survey paper accepted in Medical Image Analysis, on techniques for 3D histology reconstruction. The article, whose first author is Jonas Pichat, surveys almost 30 years of registration methods to align 2D images in a stack among themselves and also to 3D, mm-scale medical images. The article thoroughly discusses peculiarities of histological images, different types of artefacts (and ways of mitigating them), and approaches to recovering the 3D structure that is lost in the sectioning. You can find an early preprint of the manuscript under Publications.

Congrats, Jonas (et al., but mostly Jonas!) for the hard work.





Paper on drift-free mosaicking for fetoscopy

posted Jan 23, 2018, 10:31 AM by Juan Eugenio Iglesias

 
 
My coworker Marcel Tella just got a paper accepted for publication in the Journal of Medical Imaging on visual and electromagnetic data fusion for drift-free mosaicking in fetoscope. The paper describes a probabilistic framework to integrate measurements from a an electromagnetic trackers (global information, noisy) with pairwise registrations between video frames (local, accurate). You can find the paper under Publications. Thanks, Marcel (et al.) for the hard work!

Postdoc position available

posted Nov 29, 2017, 7:41 AM by Juan Eugenio Iglesias   [ updated Nov 29, 2017, 7:50 AM ]

We are looking for a postdoc to work on my ERC project "Building Next-Generation Computational Tools for High Resolution Neuroimaging Studies" at the Translational Imaging Group (University College London). If you are interested, please apply through the link at the bottom of this page.
Further details on the position can be found here below.



Two year postdoctoral position in medical image analysis (Machine Learning for 3D Histology)
Translational Imaging Group, University College London, United Kingdom.

The Translational Imaging Group (TIG) at University College London (UCL) is seeking to fill a postdoctoral position in medical image analysis within an exciting ERC Starting Grant project: "Building Next-Generation Computational Tools for High Resolution Neuroimaging Studies", awarded to Dr. J. Eugenio Iglesias. The project ultimately aims for creating an ultra-high resolution probabilistic atlas of the whole human brain using ex vivo data, as well as companion image analysis tools that apply the atlas to the automated segmentation of in vivo MRI scans.

We are seeking a talented and enthusiastic post-doctoral research associate (PDRA) to develop image analysis methods for the 3D reconstruction of the histological data in this exciting project. The PDRA will combine synthesis and segmentation techniques (Bayesian and/or deep learning based, e.g., U-Net, CycleGAN) with registration methods developed at the Translational Imaging Group. These techniques will be used to recover the 3D structure of histological samples from human brain specimens provided by the UCL Brain Bank. The resulting reconstructed data will be used to build next-generation brain atlases in later stages of the ERC project. 

This project provides an exciting opportunity for a PDRA at the interface between machine learning, medical image analysis and neuroscience, with a huge potential impact on the neuroimaging field.

The post is funded for 2 years in the first instance.

Applicants should have:

- A Ph.D. degree in computer vision, neuroimaging, or biomedical image analysis. Applicants with degrees in related areas (e.g., Engineering, Physics, Applied Mathematics) will also be considered. Applicants close to submission of their PhD thesis will be considered as well. 
- Strong mathematical and problem solving abilities.
- Expertise in programming, at least with high-level languages / packages.
- An established publication track record.
- Ideally, experience with one/some of the following: neuroimaging, deep learning, analysis of histology data, image registration.

The postdoctoral researcher would be hired at grade 7 of the UCL salary scale. The deadline for applications is December 29th. The duration of the contract would be two years, with possibility of further extension.

You can apply for the job here:



PNAS paper

posted Nov 15, 2017, 5:55 AM by Juan Eugenio Iglesias   [ updated Nov 15, 2017, 5:58 AM ]

 


It is my pleasure to announce that our paper "Sparsity enables estimation of both subcortical and cortical activity from MEG and EEG" (first author: Pavitra Krishnaswamy) has been published at PNAS. 

Localizing the sources of electromagnetic fields measured with magnetoencephalogray is very difficult, because the fields generated by deep brain structures are very weak. This paper shows that cortical and subcortical fields can be distinguished if the cortical sources are sparse.

Thanks, Pavitra (et al.) for the hard work!

Atlas of nuclei of the amygdala

posted Nov 14, 2017, 5:08 AM by Juan Eugenio Iglesias

 



Our atlas of the nuclei of the amygdala and its associated segmentation code are finally available in the development version of FreeSurfer. 

You can read about the atlas here.

The code can be found here.



1-10 of 59