Brain Decoding 2010 - 1st ICPR Workshop on Brain Decoding Pattern recognition challenges in neuroimaging Brain Decoding 2010

Introduction
While mass-univariate data processing allows to efficiently deal with large amounts of data, to maintain easy interpretation, and is widely available in existing software packages, its main drawback is that no use is made of distributed patterns of brain activity. Techniques from pattern recognition-commonly termed "multivariate pattern analysis" (MVPA) or "brain decoding" in the field of neuroimaging-have recently been recognized as a valuable extension of univariate approaches, and have already been very successful and brought new insights to brain function such as decoding of information along the visual pathway (one notable example is the detection of subconscious visual stimuli).

Many challenges of brain decoding are also relevant for other applications of pattern recognition: dealing with high-dimensional multivariate time series, dimensionality reduction, causal modelling, and properly addressing reproducibility issues within and between subjects and conditions.

Goals
In this workshop, we aim at bringing together researchers active in the field of brain decoding with those in pattern recognition. In particular, the workshop should provide a common platform for both communities to exchange knowledge and strengthen links; researchers in pattern recognition can learn about interesting and difficult problems to tackle and about new contributions to the field of pattern recognition, while researchers in neuroimaging can find out about newest trends in pattern recognition and the common challenges arising in other application domains. Of special interest is the fostering of discussions concerning interpretability of classification methods, parameters and results, a crucial point for neuroimaging where interaction with the neurosciences and medical communities is of paramount importance.