Localizing and Comparing Weight Maps Generated from Linear Kernel Machine Learning Models

Schrouff, Jessica; Crémers, Julien; Garraux, Gaëtan; Baldassarre, Luca; Mourão-Miranda, Janaı́na; Phillips, Christophe

doi:10.1109/prni.2013.40

Cited by 33 publications

(45 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To quantify the relative importance of each module, we calculated the absolute value of each weight and took the average within each module. Then, we ranked all of the modules (Schrouff et al, 2013). For the interoception prediction, the most important areas were the COTC, VAN-SSM, and COTC-SSM (Figure 5a,c); for the anxiety prediction, the most important areas were the SN-SSM, DMN-VAN, and COTC-SN (Figure 5b,c).…”

Section: Quantification Of Relative Importancementioning

confidence: 99%

Brain connection pattern under interoceptive attention state predict interoceptive intensity and subjective anxiety feeling

Shi

Wei

et al. 2018

Human Brain Mapping

View full text Add to dashboard Cite

Interoception involves the processing of a variety of different types of information ascending from the body. Accumulating evidence has indicated that interoception plays a fundamental role in cognitive and emotional processes, such as anxiety, but how different functional connectivity patterns contribute to emotions and visceral feelings during an interoceptive attention state is still unclear. In the present study, an interoceptive attention task was performed during functional magnetic resonance imaging of healthy subjects, and the participants' subjective ratings of the intensity of interoception and feelings of anxiety were recorded. Several network nodes were selected, based on previous studies, to construct task‐dependent functional connectivity patterns, which were processed by support vector regression to predict the corresponding feeling scores. The results showed that for interoception, the cingulo‐opercular task control network provided the greatest contribution, whereas the most important feature for anxiety was the connections between the sensorimotor area (SSM) and the salience network (SN). There existed four overlapping connections between the two predictions: two negative connections between the default mode network (DMN) and the SSM, one negative connection between the DMN and the SN, and one positive connection between the ventral attention network and the SN; this overlap might suggest common bodily attention processing that is involved in both interoception and anxiety. This study remediates the lack of network‐level biomarkers of interoception and provides a reference at the level of the brain for further understanding anxiety from an interoceptive perspective.

show abstract

Section: Quantification Of Relative Importancementioning

confidence: 99%

Brain connection pattern under interoceptive attention state predict interoceptive intensity and subjective anxiety feeling

Shi

Wei

et al. 2018

Human Brain Mapping

View full text Add to dashboard Cite

show abstract

“…the atlas-based local averaging method (ABLA) presented in Schrouff et al (2013a). 305 Second, an L1-MKL version of the algorithm introduced in Rakotomamonjy et al (2008) 306 and implemented in the PRoNTo toolbox (Schrouff et al, 2013b).…”

Section: Schaefer 293mentioning

confidence: 99%

Atlas-based classification algorithms for identification of informative brain regions in fMRI data

Arco

Díaz-Gutiérrez

Ramı́rez

et al. 2018

Preprint

View full text Add to dashboard Cite

Multi-voxel pattern analysis (MVPA) has been successfully applied to neuroimaging data due to its larger sensitivity compared to univariate traditional techniques. Searchlight is the most widely employed approach to assign functional value to different regions of the brain. However, its performance depends on the size of the sphere, which can overestimate the region of activation when a large sphere size is employedIn the current study, we examined the validity of two different alternatives to Searchlight: an atlas-based local averaging method (ABLA, Schrouff et al., 2013a) and a Multi-Kernel Learning (MKL, Rakotomamonjy et al., 2008) approach, in a scenario where the goal is to find the informative brain regions that support certain mental operations. These methods employ weights to measure the informativeness of a brain region and highly reduce the large computational cost that Searchlight entails. We evaluated their performance in two different scenarios where the differential BOLD activation between experimental conditions was large vs. small, and employed nine different atlases to assess the influence of diverse brain parcellations.Results show that both methods were able to localize informative regions when differences between conditions were large, demonstrating a large sensitivity and stability in the identification of regions across atlases. Moreover, the sign of the weights reported by these methods provided the directionality of univariate approaches. However, when differences were small, only ABLA localized informative regions. Thus, our results show that atlas-based methods are useful alternatives to Searchlight, but that the nature of the classification to perform should be taken into account when choosing the specific method to implement.

show abstract

“…Sensitivity, specificity, and balanced accuracy of the resulting classification solution were calculated and permutation tests based on 5,000 iterations were used to assess the level of statistical significance set at p < .05. Weight-maps and rankorders of the regional weight averages were calculated for the GM data as described in Schrouff, Cremers, et al (2013).…”

Section: Pattern Recognitionmentioning

confidence: 99%

“…side; R, right side. Area, brain region according to the automated anatomical labeling atlas(Tzourio-Mazoyer et al, 2002); Weight abs., mean of absolute weight values of included all voxels within the region(Schrouff, Cremers, et al, 2013); Weight perc., percentage of normalized weight values of this region relative to the sum of all weight values of all regions for the respective classification problem. Voxels, number of voxels included.…”

mentioning

confidence: 99%

Separating generalized anxiety disorder from major depression using clinical, hormonal, and structural MRI data: A multimodal machine learning study

et al. 2017

View full text Add to dashboard Cite

BackgroundGeneralized anxiety disorder (GAD) is difficult to recognize and hard to separate from major depression (MD) in clinical settings. Biomarkers might support diagnostic decisions. This study used machine learning on multimodal biobehavioral data from a sample of GAD, MD and healthy subjects to differentiate subjects with a disorder from healthy subjects (case‐classification) and to differentiate GAD from MD (disorder‐classification).MethodsSubjects with GAD (n = 19), MD without GAD (n = 14), and healthy comparison subjects (n = 24) were included. The sample was matched regarding age, sex, handedness and education and free of psychopharmacological medication. Binary support vector machines were used within a nested leave‐one‐out cross‐validation framework. Clinical questionnaires, cortisol release, gray matter (GM), and white matter (WM) volumes were used as input data separately and in combination.ResultsQuestionnaire data were well‐suited for case‐classification but not disorder‐classification (accuracies: 96.40%, p < .001; 56.58%, p > .22). The opposite pattern was found for imaging data (case‐classification GM/WM: 58.71%, p = .09/43.18%, p > .66; disorder‐classification GM/WM: 68.05%, p = .034/58.27%, p > .15) and for cortisol data (38.02%, p = .84; 74.60%, p = .009). All data combined achieved 90.10% accuracy (p < .001) for case‐classification and 67.46% accuracy (p = .0268) for disorder‐classification.ConclusionsIn line with previous evidence, classification of GAD was difficult using clinical questionnaire data alone. Particularly cortisol and GM volume data were able to provide incremental value for the classification of GAD. Findings suggest that neurobiological biomarkers are a useful target for further research to delineate their potential contribution to diagnostic processes.

show abstract

Localizing and Comparing Weight Maps Generated from Linear Kernel Machine Learning Models

Cited by 33 publications

References 14 publications

Brain connection pattern under interoceptive attention state predict interoceptive intensity and subjective anxiety feeling

Brain connection pattern under interoceptive attention state predict interoceptive intensity and subjective anxiety feeling

Atlas-based classification algorithms for identification of informative brain regions in fMRI data

Separating generalized anxiety disorder from major depression using clinical, hormonal, and structural MRI data: A multimodal machine learning study

Contact Info

Product

Resources

About