Model sparsity and brain pattern interpretation of classification models in neuroimaging

By Peter Mondrup Rasmussen, Lars Kai Hansen, Kristoffer Hougaard Madsen, Nathan William Churchill, Stephen C. Strother. 2012

144 citations recorded by the source corpus.

Subfields: Biophysics, Cognitive Neuroscience, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biophysics

1. John Ashburner
2. Christos Davatzikos
3. Gaël Varoquaux

In Cognitive Neuroscience

1. Gaël Varoquaux
2. John Ashburner
3. Bertrand Thirion
4. Christos Davatzikos
5. Yannick Schwartz
6. Christophe L. M. Phillips
7. Janaina Mourão Miranda
8. Maria João Duarte Rosa

In Radiology, Nuclear Medicine and Imaging

1. Christos Davatzikos
2. Aristeidis Sotiras
3. Carlton Chu
4. John Ashburner
5. Jonas Richiardi
6. Gaël Varoquaux
7. Bertrand Thirion
8. Christophe L. M. Phillips

Citations in the corpus

• PRoNTo: Pattern Recognition for Neuroimaging Toolbox (2013, background): “Rasmussen et al. (2012) introduce an interesting discussion regarding the interpretability and generalization of the model weights.”
• Generative Discriminative Models for Multivariate Inference and Statistical Mapping in Medical Imaging (2018, methodology): “This results in improved sensitivity and numerous approaches have been proposed to efficiently obtain meaningful multivariate brain patterns [10, 14, 13, 4, 7, 6].”
• Analytic estimation of statistical significance maps for support vector machine based multi-variate image analysis and classification (2013, unknown): “The authors of Rasmussen et al. (2012) have previously noted this for neuroimaging data.”
• Assessing and tuning brain decoders: Cross‐validation, caveats, and guidelines (2016, methodology): “However, logistic re-gressions (Log-Reg) are also often used [7, 50, 52, 57, 60].”
• Multi‐center machine learning in imaging psychiatry: A meta‐model approach (2017, unknown): “On the other hand, studies have shown that neuroimaging data are not always fully linearly separable (Rasmussen et al., 2012), which is reflected by the influence of parameter C in SVM models.”
• Relevant Feature Set Estimation with a Knock-out Strategy and Random Forests (2015, unknown): “There have been attempts to tackle some of the issues by Meinshausen and Bühlmann (2010), Rasmussen et al. (2012), Rondina et al. (2014), Rondina et al. (2013), Haufe et al. (2014), and Yourganov et al. (2014), but none of these algorithms could ameliorate all problems.”