[Neuroinfo] Call for Papers: Special Session on Computational Methods for Neuroimaging Analysis, 6-8 September, Portugal
Tiago Azevedo
tiago.azevedo at cst.cam.ac.uk
Fri May 18 13:13:24 CEST 2018
Dear all,
Below find the call for papers for the special session "Computational
Methods for Neuroimaging Analysis". It will be be held during the 15th
International Conference on Computational Intelligence methods for
Bioinformatics and Biostatistics, at Caparica, Portugal, from 6 to 8
September.
**Important Dates**
Paper submission deadline: 10 June 2018
Acceptance notification: 9 July 2018
Author registration due: 20 July 2018
Camera ready due: 29 July 2018
Conference: 6-8 September 2018
**Link**
https://eventos.fct.unl.pt/cibb2018/pages/special-sessions
**Aim and scope**
There is an increasing need for the application of machine learning (ML)
techniques which can perform image processing operations such as
segmentation, coregistration, classification and dimensionality
reduction in the field of neuroimaging. Although the manual approach
often remains the golden standard in some tasks (like segmentation), ML
can be utilised to automate and facilitate the work of researchers and
clinicians. Frequently used techniques include support vector machines
(SVMs) for classification problems, graph-based methods for brain
network analysis and recently artificial neural networks (ANNs).
Deep ANNs, i.e. deep learning, have proved to be very successful in
computer vision tasks owing to the ability to automatically extract
hierarchical descriptive features from input images. It has also been
used in the medical and neuroimaging domains for automatic disease
diagnosis, tissue segmentation and even synthetic image generation. The
issue, however, is the relative sample paucity in typical neuroimaging
datasets which leads to poor generalisation considering the high number
of parameters employed in typical deep neural networks. Consequently,
parameter- efficient design paradigms ought to be created.
Another approach to investigate degeneration is the study and mapping of
the neural connections in the brain known as the connectome. The
connectome can be seen as a matrix representing all possible pairwise
connections between different neural areas. Researchers study both the
structural and functional connectivity in order to understand important
brain patterns, such as how the connectome impacts the dynamics of
disease spreading, ageing and learning.
Topics of interest includes but are not limited to:
• Machine learning techniques for segmentation, coregistration,
classification or dimensionality reduction of neuroimages
• Deep learning for neuroimaging analysis
• Brain network analysis
• Applications of graph theory to MRI and fMRI data
• Applications of machine learning methodologies for neurodegenerative
disease studies
• Computational modelling and analysis of neuroimaging
• Methods of analysis for structural or functional connectivity
• Development of new neuroimaging tools
**Session chairs**
Tiago Azevedo, University of Cambridge, UK
Giovanna Maria Dimitri, University of Cambridge, UK
Pietro Liò, University of Cambridge, UK
Angela Serra, University of Salerno, Italy
Simeon Spasov, University of Cambridge, UK
More information about the Neuroinfo
mailing list