Visualization of nonlinear kernel models in neuroimaging by sensitivity maps

Rasmussen, Peter Mondrup; Madsen, Kristoffer Hougaard; Lund, Torben Ellegaard; Hansen, Lars Kai

doi:10.1016/j.neuroimage.2010.12.035

Cited by 59 publications

(73 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In imaging this corresponds to a specific voxel. While the weight map itself has been used for interpreting SVM models Rasmussen et al (2011) Guyon et al (2002), it has been noted that using SVM weights can assign relatively low weights to significant features and relatively larger weights to irrelevant features Hardin et al (2004) Cuingnet et al (2011) Gaonkar and Davatzikos (2013). We have also documented this behavior in figure 4 and the associated experiment.…”

Section: Methodssupporting

confidence: 62%

“…Further, this section would perhaps be incomplete without mentioning the work of Cuingnet et al (2011) and Rasmussen et al (2011) both of which influenced the drafting of this document. Specifically, Cuingnet et al (2011) introduced us to the concept of SVM based permutation testing.…”

Section: Introductionmentioning

confidence: 99%

“…Specifically, Cuingnet et al (2011) introduced us to the concept of SVM based permutation testing. The work by Rasmussen et al (2011) generalizes the concept of SVM weight based visualization to kernel based non-linear classification frameworks. While there exist certain disadvantages to using SVM weights directly (see Cuingnet et al (2011)), the work by Rasmussen et al (2011) provides a roadmap for future development of this method.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Interpreting support vector machine models for multivariate group wise analysis in neuroimaging

Gaonkar¹,

Shinohara

Davatzikos³

2015

Medical Image Analysis

View full text Add to dashboard Cite

Machine learning based classification algorithms like support vector machines (SVMs) have shown great promise for turning a high dimensional neuroimaging data into clinically useful decision criteria. However, tracing imaging based patterns that contribute significantly to classifier decisions remains an open problem. This is an issue of critical importance in imaging studies seeking to determine which anatomical or physiological imaging features contribute to the classifier’s decision, thereby allowing users to critically evaluate the findings of such machine learning methods and to understand disease mechanisms. The majority of published work addresses the question of statistical inference for support vector classification using permutation tests based on SVM weight vectors. Such permutation testing ignores the SVM margin, which is critical in SVM theory. In this work we emphasize the use of a statistic that explicitly accounts for the SVM margin and show that the null distributions associated with this statistic are asymptotically normal. Further, our experiments show that this statistic is a lot less conservative as compared to weight based permutation tests and yet specific enough to tease out multivariate patterns in the data. Thus, we can better understand the multivariate patterns that the SVM uses for neuroimaging based classification.

show abstract

Section: Methodssupporting

confidence: 62%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Interpreting support vector machine models for multivariate group wise analysis in neuroimaging

Gaonkar¹,

Shinohara

Davatzikos³

2015

Medical Image Analysis

View full text Add to dashboard Cite

show abstract

“…The usefulness of such nonlinear classifiers for neuroimaging applications is the subject of an on-going debate in the litterature (see the introduction of [79] for a summary). The appeal of linear classifiers for fMRI applications is mainly twofold: i) they facilitate the interpretation of the classification results, for instance thanks to the ability to directly visualize weight maps when working with linear SVM [80]; and ii) despite their simplicity, their performances always are amongst the highest [81].…”

Section: Discussionmentioning

confidence: 99%

“…The appeal of linear classifiers for fMRI applications is mainly twofold: i) they facilitate the interpretation of the classification results, for instance thanks to the ability to directly visualize weight maps when working with linear SVM [80]; and ii) despite their simplicity, their performances always are amongst the highest [81]. Regarding the first point, [79] offers a solution to ease the interpretation of the results given by nonlinear classifiers by visualizing sensitivity maps. As for the second point, our study is a new example where a nonlinear method clearly outperforms linear classifiers.…”

Section: Discussionmentioning

confidence: 99%

Graph-Based Inter-Subject Pattern Analysis of fMRI Data

et al. 2014

View full text Add to dashboard Cite

In brain imaging, solving learning problems in multi-subjects settings is difficult because of the differences that exist across individuals. Here we introduce a novel classification framework based on group-invariant graphical representations, allowing to overcome the inter-subject variability present in functional magnetic resonance imaging (fMRI) data and to perform multivariate pattern analysis across subjects. Our contribution is twofold: first, we propose an unsupervised representation learning scheme that encodes all relevant characteristics of distributed fMRI patterns into attributed graphs; second, we introduce a custom-designed graph kernel that exploits all these characteristics and makes it possible to perform supervised learning (here, classification) directly in graph space. The well-foundedness of our technique and the robustness of the performance to the parameter setting are demonstrated through inter-subject classification experiments conducted on both artificial data and a real fMRI experiment aimed at characterizing local cortical representations. Our results show that our framework produces accurate inter-subject predictions and that it outperforms a wide range of state-of-the-art vector- and parcel-based classification methods. Moreover, the genericity of our method makes it is easily adaptable to a wide range of potential applications. The dataset used in this study and an implementation of our framework are available at http://dx.doi.org/10.6084/m9.figshare.1086317.

show abstract

Multivariate linear regression of high‐dimensional fMRI data with multiple target variables

Valente

Castellanos

Vanacore

et al. 2013

Human Brain Mapping

View full text Add to dashboard Cite

Multivariate regression is increasingly used to study the relation between fMRI spatial activation patterns and experimental stimuli or behavioral ratings. With linear models, informative brain locations are identified by mapping the model coefficients. This is a central aspect in neuroimaging, as it provides the sought-after link between the activity of neuronal populations and subject's perception, cognition or behavior. Here, we show that mapping of informative brain locations using multivariate linear regression (MLR) may lead to incorrect conclusions and interpretations. MLR algorithms for high dimensional data are designed to deal with targets (stimuli or behavioral ratings, in fMRI) separately, and the predictive map of a model integrates information deriving from both neural activity patterns and experimental design. Not accounting explicitly for the presence of other targets whose associated activity spatially overlaps with the one of interest may lead to predictive maps of troublesome interpretation. We propose a new model that can correctly identify the spatial patterns associated with a target while achieving good generalization. For each target, the training is based on an augmented dataset, which includes all remaining targets. The estimation on such datasets produces both maps and interaction coefficients, which are then used to generalize. The proposed formulation is independent of the regression algorithm employed. We validate this model on simulated fMRI data and on a publicly available dataset. Results indicate that our method achieves high spatial sensitivity and good generalization and that it helps disentangle specific neural effects from interaction with predictive maps associated with other targets.

show abstract

Visualization of nonlinear kernel models in neuroimaging by sensitivity maps

Cited by 59 publications

References 50 publications

Interpreting support vector machine models for multivariate group wise analysis in neuroimaging

Interpreting support vector machine models for multivariate group wise analysis in neuroimaging

Graph-Based Inter-Subject Pattern Analysis of fMRI Data

Multivariate linear regression of high‐dimensional fMRI data with multiple target variables

Contact Info

Product

Resources

About