Learning Connectivity Patterns via Graph Kernels for fMRI-Based Depression Diagnostics

It has long been known that patients with depression exhibit abnormal brain functional connectivity patterns, that are often studied from a graph-theoretic perspective. However, while certain simpler graph features have been examined, little has been done in the direction of advanced feature learning methodologies such as network embeddings. Our work aims to extend the understanding of importance of graph-based features for medical applications by evaluating the recently proposed anonymous walk embeddings (AWE) in difficult depression classification problems. For two challenging datasets, we obtain performance gains and investigate the learned vector representations. Our results indicate that using AWE-based features is a promising new direction for medical applications.

[1]  Evgeny Burnaev,et al.  Model selection for anomaly detection , 2015, International Conference on Machine Vision.

[2]  Wayne Katon,et al.  Depression and pain. , 2009, Frontiers in bioscience.

[3]  Xin Wang,et al.  Depression Disorder Classification of fMRI Data Using Sparse Low-Rank Functional Brain Network and Graph-Based Features , 2017, Comput. Math. Methods Medicine.

[4]  Quoc V. Le,et al.  Distributed Representations of Sentences and Documents , 2014, ICML.

[5]  Sergey Ivanov,et al.  Anonymous Walk Embeddings , 2018, ICML.

[6]  Anthony Widjaja,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2003, IEEE Transactions on Neural Networks.

[7]  Bruce Hermann,et al.  Consensus statement: The evaluation and treatment of people with epilepsy and affective disorders , 2008, Epilepsy & Behavior.

[8]  Yong-Ku Kim,et al.  Application of machine learning classification for structural brain MRI in mood disorders: Critical review from a clinical perspective , 2018, Progress in Neuro-Psychopharmacology and Biological Psychiatry.

[9]  H. Aizenstein,et al.  Studying depression using imaging and machine learning methods , 2015, NeuroImage: Clinical.

[10]  Evgeny Burnaev,et al.  MRI-Based Diagnostics of Depression Concomitant with Epilepsy: In Search of the Potential Biomarkers , 2018, 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA).

[11]  Benson Mwangi,et al.  A Review of Feature Reduction Techniques in Neuroimaging , 2013, Neuroinformatics.

[12]  Andrew T. Drysdale,et al.  Resting-state connectivity biomarkers define neurophysiological subtypes of depression , 2016, Nature Medicine.

[13]  Kai Li,et al.  Computational approaches to fMRI analysis , 2017, Nature Neuroscience.

[14]  Tapio Salakoski,et al.  A comparison of AUC estimators in small-sample studies , 2009, MLSB.

[15]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[16]  Evgeny Burnaev,et al.  Influence of resampling on accuracy of imbalanced classification , 2015, International Conference on Machine Vision.

[17]  Evgeny Burnaev,et al.  On an iterative algorithm for calculating weighted principal components , 2015 .

[18]  Kurt Mehlhorn,et al.  Weisfeiler-Lehman Graph Kernels , 2011, J. Mach. Learn. Res..

[19]  S. Costafreda,et al.  Neuroimaging-Based Biomarkers in Psychiatry: Clinical Opportunities of a Paradigm Shift , 2013, Canadian journal of psychiatry. Revue canadienne de psychiatrie.

[20]  J. P. Hamilton,et al.  Neural systems approaches to understanding major depressive disorder: An intrinsic functional organization perspective , 2013, Neurobiology of Disease.

[21]  Y. Benjamini,et al.  Resampling-based false discovery rate controlling multiple test procedures for correlated test statistics , 1999 .

[22]  David Haussler,et al.  Convolution kernels on discrete structures , 1999 .

[23]  Mathias Niepert,et al.  Learning Convolutional Neural Networks for Graphs , 2016, ICML.

[24]  Evgeny Burnaev,et al.  The influence of parameter initialization on the training time and accuracy of a nonlinear regression model , 2016 .

[25]  Jean-Baptiste Poline,et al.  Brain covariance selection: better individual functional connectivity models using population prior , 2010, NIPS.

[26]  W. Hauser,et al.  Major depression is a risk factor for seizures in older adults , 2000, Annals of neurology.

[27]  Kevin Chen-Chuan Chang,et al.  A Comprehensive Survey of Graph Embedding: Problems, Techniques, and Applications , 2017, IEEE Transactions on Knowledge and Data Engineering.

[28]  László Babai,et al.  Graph isomorphism in quasipolynomial time [extended abstract] , 2016, STOC.

[29]  Evgeny V. Burnaev,et al.  On a method for constructing ensembles of regression models , 2013, Autom. Remote. Control..

[30]  Pinar Yanardag,et al.  Deep Graph Kernels , 2015, KDD.

[31]  Michalis Vazirgiannis,et al.  Classifying Graphs as Images with Convolutional Neural Networks , 2017, ArXiv.

[32]  Christopher Dowrick,et al.  Self-Rated Health and Long-Term Prognosis of Depression , 2014, The Annals of Family Medicine.

[33]  Yoshua Bengio,et al.  A Neural Probabilistic Language Model , 2003, J. Mach. Learn. Res..

[34]  Evgeny Burnaev,et al.  Pattern Recognition Pipeline for Neuroimaging Data , 2018, ANNPR.

[35]  Jonathan Masci,et al.  Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[36]  Evgeny Burnaev,et al.  One-Class SVM with Privileged Information and Its Application to Malware Detection , 2016, 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW).

[37]  Wei Lu,et al.  Deep Neural Networks for Learning Graph Representations , 2016, AAAI.

[38]  Azeem Majeed,et al.  The Epidemiology of the Comorbidity of Epilepsy in the General Population , 2004, Epilepsia.

[39]  S. V. N. Vishwanathan,et al.  Graph kernels , 2007 .

[40]  Silvio Micali,et al.  Reconstructing Markov processes from independent and anonymous experiments , 2016, Discret. Appl. Math..