NIPS 2011 workshop on machine learning and interpretation in neuroimaging

Download printable schedule and abstracts


Modern multivariate statistical methods have been increasingly applied to various problems in neuroimaging, including “mind reading”, “brain mapping”, clinical diagnosis and prognosis. Multivariate pattern analysis (MVPA) is a promising machine-learning approach for discovering  complex relationships between high-dimensional signals (e.g., brain images) and variables of interest (e.g., external stimuli and/or  brain's cognitive states). Modern multivariate regularization approaches can overcome the curse of dimensionality and produce highly predictive models even in  high-dimensional, small-sample scenarios  typical in neuroimaging (e.g., 10 to 100 thousands of voxels  and just a few hundreds of samples).

However, despite the rapidly growing number of neuroimaging applications in machine learning, its impact on how theories of brain function are construed has received little consideration. Accordingly, machine-learning techniques are frequently met with skepticism in the domain of cognitive neuroscience. In this workshop, we intend to investigate the implications that follow from adopting machine-learning methods for studying brain function. In particular, this concerns the question how these methods may be used to represent cognitive states, and what ramifications this has for consequent theories of cognition. Besides providing a rationale for the use of machine-learning methods in studying brain function, a further goal of this workshop is to identify shortcomings of state-of-the-art approaches and initiate research efforts that increase the impact of machine learning on cognitive neuroscience.

Decoding higher cognition and interpreting the behavior of associated classifiers can pose unique challenges, as these psychological states are complex, fast-changing and often ill-defined. For instance, speech is received at 3-4 words a second; acoustic, semantic and syntactic processing occur in parallel; and the form of underlying representations (sentence structures, conceptual descriptions) remains controversial. ML techniques are required that can take advantage of patterns that are temporally and spatially distributed, but coordinated in their activity. And different recording modalities have distinctive advantages: fMRI provides millimeter-level localization in the brain but poor temporal resolution, while EEG and MEG have millisecond temporal resolution at the cost of spatial resolution. Ideally machine learning methods would be able to meaningfully combine complementary information from these different neuroimaging techniques, and reveal latent dimensions in neural activity, while still being capable of disentangling tightly linked and confounded sub-processes.

Moreover, from the machine learning perspective, neuroimaging is a  rich source of challenging problems that can facilitate development of novel approaches. For example, feature extraction and feature selection approaches  become particularly important  in neuroimaging, since the primary objective  is to gain a scientific insight rather than simply learn a ``black-box'' predictor. However, unlike some other applications  where the set features might be quite well-explored and established by now, neuroimaging is a domain where  a machine-learning researcher cannot simply "ask domain experts what features should be used", since this is essentially the question domain experts themselves are trying to figure out.  While the current  neuroscientific knowledge can  guide the definition of  specialized  'brain areas', more complex patterns of brain activity, such as spatio-temporal patterns, functional network patterns, and other multivariate dependencies  remain to be discovered mainly via statistical analysis.

Open questions

The list of open questions of interest to the workshop includes, but is not limited to the following:

  • How can we interpret results of multivariate models in a neuroscientific context? 
  • How suitable are MVPA and inference methods for brain mapping?
  • How can we assess the specificity and sensitivity?
  • What is the role of decoding vs. embedded or separate feature selection?
  • How can we use these approaches for a flexible and useful representation of neuroimaging data?
  • What can we accomplish with generative vs. discriminative modelling?
  • How can ML techniques help us in modeling higher cognitive processes (e.g. reasoning, communication, knowledge representation)?
  • How can we disentangle confounded processes and representations?
  • How do we combine the data from different  recording modalities (e.g. fMRI, EEG, structural MRI, DTI, MEG, NIRS, EcOG, single cell recordings, etc.)?

The workshop

In this two-day workshop we will explore perspectives and novel methodology at the interface of Machine Learning, Inference, Neuroimaging and Neuroscience. We aim to bring researchers from machine learning and neuroscience community together, in order to discuss open questions, identify the core points for a number of the controversial issues, and  eventually propose approaches to solving those issues.

The workshop will be structured around 4 main topics:

  • Machine learning and pattern recognition methodology
  • Interpretable decoding of higher cognitive states from neural data
  • Causal inference in neuroimaging
  • Linking machine learning, neuroimaging and neuroscience

Each session will be opened by 2-3 invited talks, and an in depth discussion. This will be followed by original contributions. Original contributions will also be presented and discussed during a poster session. Each day of the workshop will end with a panel discussion, during which we will address specific questions, and invited speakers will open each segment with a brief presentation of their opinion.

This workshop is part of the PASCAL2 Thematic Programme on Cognitive Inference and Neuroimaging (http://mlin.kyb.tuebingen.mpg.de/).