Publications of Marie Wehenkel
Bookmark and Share    
Full Text
See detailCharacterization of neurodegenerative diseases with tree ensemble methods: the case of Alzheimer's disease
Wehenkel, Marie ULiege

Doctoral thesis (2018)

For the last decade, the neuroscience field has observed the emergence of machine learning methods for the analysis of neuroimaging data. Unlike univariate methods that consider voxels one per one, these ... [more ▼]

For the last decade, the neuroscience field has observed the emergence of machine learning methods for the analysis of neuroimaging data. Unlike univariate methods that consider voxels one per one, these techniques analyse relationships between several voxels and are able to detect multivariate patterns. In the context of neurodegenerative diseases, such as Alzheimer’s disease (AD), they can be used to design a diagnosis system and to find in neuroimages the patterns responsible for the disease. The context of the work presented here is thus the field of pattern recognition with neuroimaging. Our objective is to explore the possibilities that tree ensemble methods, such as Random Forests, offer in this domain in general, and in particular in the context of AD research. These methods suit very well the needs of this domain, as they combine very good predictive performances and provide interpretable results in the form of variable importance scores. Our contributions include both methodological developments around tree ensemble methods and applications of these methods on real datasets. The methodological part of the thesis focuses on the analysis and the improvement of Random Forests variable importances for neuroimaging problems. Typical datasets in this domain are of very high dimensionality (hundreds of thousands of voxels) and contain comparatively very few samples (tens or hundreds of patients). Our first contribution is a theoretical and empirical analysis of how importance scores behave in such extreme settings, depending on the method parameters. We then propose several improvements of importance scores in such settings that take advantage of either the spatial structure between the features or a pre-defined partitioning of these features into groups. Finally, we address an issue with Random Forests importances, which is to find a threshold between truly relevant and irrelevant variables. For this purpose, we adapt several statistical methods proposed in the bioinformatics literature. These methods are extended to compute a statistical score for groups of features instead of individual features. This adaptation at the group level has been raised from our expectation to find groups of voxels explaining a disease instead of isolated voxels. We show that working at the group level leads to a higher statistical power than working at the feature level. The approach is applied on a real dataset for the prognosis of AD, where it is shown to highlight brain regions that are consistent with results in the literature. In the second part of the thesis, we show different applications of Random Forests for AD research. First, we use tree-based ensemble methods in order to clinically characterize two different metabolic profiles observed in PET scans of AD patients. Second, we carry out an empirical comparison that shows that Random Forests are competitive with linear methods, in terms of accuracy and interpretability, on different real datasets related to three research questions about AD: the diagnosis of demented patients, the prognosis of mild cognitively impaired (MCI) patients, and the differentiation of MCI and AD patients. [less ▲]

Detailed reference viewed: 386 (46 ULiège)
Full Text
See detailRandom Forests based group importance scores and their statistical interpretation: application for Alzheimer’s disease
Wehenkel, Marie ULiege; Sutera, Antonio ULiege; Bastin, Christine ULiege et al

in Frontiers in Neuroscience (2018), 12

Machine learning approaches have been increasingly used in the neuroimaging field for the design of computer-aided diagnosis systems. In this paper, we focus on the ability of these methods to provide ... [more ▼]

Machine learning approaches have been increasingly used in the neuroimaging field for the design of computer-aided diagnosis systems. In this paper, we focus on the ability of these methods to provide interpretable information about the brain regions that are the most informative about the disease or condition of interest. In particular, we investigate the benefit of group-based, instead of voxel-based, analyses in the context of Random forests. Assuming a prior division of the voxels into non overlapping groups (defined by an atlas), we propose several procedures to derive group importances from individual voxel importances derived from random forests models. We then adapt several permutation schemes to turn group importance scores into more interpretable statistical scores that allow to determine the truly relevant groups in the importance rankings. The good behavior of these methods is first assessed on artificial datasets. Then, they are applied on our own dataset of FDG-PET scans to identify the brain regions involved in the prognosis of Alzheimer's disease. [less ▲]

Detailed reference viewed: 44 (13 ULiège)
Full Text
See detailComputer Aided Diagnosis System Based on Random Forests for the Prognosis of Alzheimer’s Disease
Wehenkel, Marie ULiege; Bastin, Christine ULiege; Geurts, Pierre ULiege et al

in 1st HBP Student Conference - Transdisciplinary Research Linking Neuroscience, Brain Medicine and Computer Science (2018, April)

In this abstract, we propose an original CAD system consisting in the combination of brain parcelling, ensemble of trees methods, and selection of (groups of) features using the importance scores embedded ... [more ▼]

In this abstract, we propose an original CAD system consisting in the combination of brain parcelling, ensemble of trees methods, and selection of (groups of) features using the importance scores embedded in tree-based methods. Indeed, on top of their ease of use and accuracy without ad hoc parameter tuning, tree ensemble methods such as random forests (RF) (Breiman, 2001) or extremely randomized trees (ET) (Geurts et al., 2006) provide interpretable results in the form of feature importance scores. We also compare the performance and interpretability of our proposed method to standard RF and ET approaches, without feature selection, and to multiple kernel learning (MKL). The latter was shown to be an efficient method notably capable of dealing with anatomically defined regions of the brain by the use of multiple kernels. [less ▲]

Detailed reference viewed: 32 (5 ULiège)
Full Text
See detailTree Ensemble Methods and Parcelling to Identify Brain Areas Related to Alzheimer’s Disease
Wehenkel, Marie ULiege; Bastin, Christine ULiege; Phillips, Christophe ULiege et al

in 2017 International Workshop on Pattern Recognition in Neuroimaging (PRNI), proceedings (2017, June)

Detailed reference viewed: 79 (22 ULiège)
Full Text
See detailAccuracy and interpretability, tree-based machine learning approaches in neuroimaging
Wehenkel, Marie ULiege

Poster (2016, January 25)

Detailed reference viewed: 29 (8 ULiège)
Full Text
See detailForward Stagewise Additive Modeling for Computer Aided Diagnosis (CAD) Systems
Wehenkel, Marie ULiege

Poster (2015, November 26)

Detailed reference viewed: 36 (7 ULiège)
Full Text
See detailTree Ensemble Methods for Computer Aided Diagnosis (CAD) Systems : Application to neurodegenerative diseases
Wehenkel, Marie ULiege

in Human Brain Project : 2nd HBP Education Workshop : Future Medicine : Medical Intelligence for Brain Diseases : 15th-18th March 2015, CHUV, Lausanne, Switzerland : Abstract Collection Students (2015, March)

Detailed reference viewed: 45 (12 ULiège)
Full Text
See detailCellular regulation mechanisms : case study of up and down states in the Purkinje cell
Wehenkel, Marie ULiege

Master's dissertation (2014)

In 1963, Hodgkin and Huxley obtained the Nobel Prize to have shown that the electrical activity of a neuron could be modelled by an electrical RC circuit containing non-linear conductances. This discovery ... [more ▼]

In 1963, Hodgkin and Huxley obtained the Nobel Prize to have shown that the electrical activity of a neuron could be modelled by an electrical RC circuit containing non-linear conductances. This discovery made it possible to reproduce the electrical behaviour of neurons with a level of detail that has steadily increased over the last fifty years as new quantitative knowledge became available about the specific ionic currents that regulate the activity of a given neuron. But models with too many details are often non-robust and too complex for analysis. As control engineers need simplified models for control design, experimental neurophysiologists are in need of models that are amenable to sensitivity and robustness analysis, beyond the mere simulation of a given neuronal behaviour recorded experimentally. The Purkinje cell has been studied for over hundred years because its large dendritic tree enables to recognize it easily with a microscope. This neuron exhibits a bistability between a stable hyperpolarized down-state and a stable depolarized spiking state. It is one of the first discovered neurons, however its electrical behaviour is not well understood so far. The principal question of the thesis is to model the electrophysiology of the Purkinje cell to advance the understanding of its regulation mechanisms. More particularly, the objective of the thesis is to explore recent work about the role of the calcium current in neuronal excitability as a possible mechanism underlying the bistability observed in the Purkinje cell. The electrical activity of the Purkinje cell is reproduced in this thesis thanks to a reduced physiological model which can be seen as an intermediate between a detailed model with dendritic compartments and an abstract model of bistability. This novel model is the main contribution of the thesis. Its main ingredients are on the one hand a fast sodium current and a slow potassium restorative current whose particular kinetics account for the up-state excitability, and on the other hand a slow regenerative calcium current and an ultraslow calcium-dependent potassium current for bistability. The proposed model suggests several implications. First, a complex compartmental model seems unnecessary to reproduce the electrophysiology of the cell, although the profuse dendrites are an important characteristic of the Purkinje neuron. Secondly, the Purkinje neuron appears to be regulated by the same mechanisms as other bistable neurons such as the thalamocortical (TC) or subthalamic nucleus (STN) neurons. Its behaviour depends on the same feedback mechanisms (a fast regenerative sodium current, a slow restorative potassium current and a slow regenerative calcium current), event though the temporal signature is markedly different because of the specific channel kinetics primarily of the slow potassium current. Finally this novel model makes the Purkinje cell modelling amenable to robustness and modulation studies, as recently shown for similar neurons. [less ▲]

Detailed reference viewed: 147 (31 ULiège)