Christian Desrosiers

A Comprehensive Survey of Neighborhood-Based Recommendation Methods

Among collaborative recommendation approaches, methods based on nearest-neighbors still enjoy a huge amount of popularity, due to their simplicity, their efficiency, and their ability to produce accurate and personalized recommendations. This chapter presents a comprehensive survey of neighborhood-based methods for the item recommendation problem. In particular, the main benefits of such methods, as well as their principal characteristics, are described. Furthermore, this document addresses the essential decisions that are required while implementing a neighborhood-based recommender system, and gives practical information on how to make such decisions. Finally, the problems of sparsity and limited coverage, often observed in large commercial recommender systems, are discussed, and a few solutions to overcome these problems are presented.

Within-network classification using local structure similarity

Within-network classification, where the goal is to classify the nodes of a partly labeled network, is a semi-supervised learning problem that has applications in several important domains like image processing, the classification of documents, and the detection of malicious activities. While most methods for this problem infer the missing labels collectively based on the hypothesis that linked or nearby nodes are likely to have the same labels, there are many types of networks for which this assumption fails, e.g., molecular graphs, trading networks, etc. In this paper, we present a collective classification method, based on relaxation labeling, that classifies entities of a network using their local structure. This method uses a marginalized similarity kernel that compares the local structure of two nodes with random walks in the network. Through experimentation on different datasets, we show our method to be more accurate than several state-of-the-art approaches for this problem.

3D fully convolutional networks for subcortical segmentation in MRI: A large-scale study

ABSTRACT This study investigates a 3D and fully convolutional neural network (CNN) for subcortical brain structure segmentation in MRI. 3D CNN architectures have been generally avoided due to their computational and memory requirements during inference. We address the problem via small kernels, allowing deeper architectures. We further model both local and global context by embedding intermediate‐layer outputs in the final prediction, which encourages consistency between features extracted at different scales and embeds fine‐grained information directly in the segmentation process. Our model is efficiently trained end‐to‐end on a graphics processing unit (GPU), in a single stage, exploiting the dense inference capabilities of fully CNNs. We performed comprehensive experiments over two publicly available datasets. First, we demonstrate a state‐of‐the‐art performance on the ISBR dataset. Then, we report a large‐scale multi‐site evaluation over 1112 unregistered subject datasets acquired from 17 different sites (ABIDE dataset), with ages ranging from 7 to 64 years, showing that our method is robust to various acquisition protocols, demographics and clinical factors. Our method yielded segmentations that are highly consistent with a standard atlas‐based approach, while running in a fraction of the time needed by atlas‐based methods and avoiding registration/normalization steps. This makes it convenient for massive multi‐site neuroanatomical imaging studies. To the best of our knowledge, our work is the first to study subcortical structure segmentation on such large‐scale and heterogeneous data.

Vesselwalker: Coronary arteries segmentation using random walks and hessian-based vesselness filter

The segmentation of vascular structures from 2D X-ray angiographies is an important step for vessel measurement, diagnosis and treatment planning. Segmentation of such structures can be challenging due to the vessel appearance and topology. In this paper, we propose a novel interactive method to segment vascular structures by combining Hessian-based vesselness information and the random walk formulation, in which manually selected seed points can be used to refine the segmentation result. The proposed method was tested on coronary arteries angiograms and has shown to be more accurate than an active contour-based method or the Random Walker algorithm, with a mean AUC of 97.2%.

Multi texture analysis of colorectal cancer continuum using multispectral imagery

Purpose This paper proposes to characterize the continuum of colorectal cancer (CRC) using multiple texture features extracted from multispectral optical microscopy images. Three types of pathological tissues (PT) are considered: benign hyperplasia, intraepithelial neoplasia and carcinoma. Materials and Methods In the proposed approach, the region of interest containing PT is first extracted from multispectral images using active contour segmentation. This region is then encoded using texture features based on the Laplacian-of-Gaussian (LoG) filter, discrete wavelets (DW) and gray level co-occurrence matrices (GLCM). To assess the significance of textural differences between PT types, a statistical analysis based on the Kruskal-Wallis test is performed. The usefulness of texture features is then evaluated quantitatively in terms of their ability to predict PT types using various classifier models. Results Preliminary results show significant texture differences between PT types, for all texture features (p-value < 0.01). Individually, GLCM texture features outperform LoG and DW features in terms of PT type prediction. However, a higher performance can be achieved by combining all texture features, resulting in a mean classification accuracy of 98.92%, sensitivity of 98.12%, and specificity of 99.67%. Conclusions These results demonstrate the efficiency and effectiveness of combining multiple texture features for characterizing the continuum of CRC and discriminating between pathological tissues in multispectral images.

A novel approach to compute similarities and its application to item recommendation

Several key applications like recommender systems deal with data in the form of ratings made by users on items. In such applications, one of the most crucial tasks is to find users that share common interests, or items with similar characteristics. Assessing the similarity between users or items has several valuable uses, among which are the recommendation of new items, the discovery of groups of like-minded individuals, and the automated categorization of items. It has been recognized that popular methods to compute similarities, based on correlation, are not suitable for this task when the rating data is sparse. This paper presents a novel approach, based on the SimRank algorithm, to compute similarity values when ratings are limited. Unlike correlation-based methods, which only consider user ratings for common items, this approach uses all the available ratings, allowing it to compute meaningful similarities. To evaluate the usefulness of this approach, we test it on the problem of predicting the ratings of users for movies and jokes.

Efficient algorithms for finding critical subgraphs

This paper presents algorithms to find vertex-critical and edge-critical subgraphs in a given graph G, and demonstrates how these critical subgraphs can be used to determine the chromatic number of G. Computational experiments are reported on random and DIMACS benchmark graphs to compare the proposed algorithms, as well as to find lower bounds on the chromatic number of these graphs. We improve the best known lower bound for some of these graphs, and we are even able to determine the chromatic number of some graphs for which only bounds were known.

Classifications of multispectral colorectal cancer tissues using convolution neural network

Background: Colorectal cancer (CRC) is the third most common cancer among men and women. Its diagnosis in early stages, typically done through the analysis of colon biopsy images, can greatly improve the chances of a successful treatment. This paper proposes to use convolution neural networks (CNNs) to predict three tissue types related to the progression of CRC: benign hyperplasia (BH), intraepithelial neoplasia (IN), and carcinoma (Ca). Methods: Multispectral biopsy images of thirty CRC patients were retrospectively analyzed. Images of tissue samples were divided into three groups, based on their type (10 BH, 10 IN, and 10 Ca). An active contour model was used to segment image regions containing pathological tissues. Tissue samples were classified using a CNN containing convolution, max-pooling, and fully-connected layers. Available tissue samples were split into a training set, for learning the CNN parameters, and test set, for evaluating its performance. Results: An accuracy of 99.17% was obtained from segmented image regions, outperforming existing approaches based on traditional feature extraction, and classification techniques. Conclusions: Experimental results demonstrate the effectiveness of CNN for the classification of CRC tissue types, in particular when using presegmented regions of interest.

Image denoising based on sparse representation and gradient histogram

Various image priors, such as sparsity prior, non-local self-similarity prior and gradient histogram prior, have been widely used for noise removal, while preserving the image texture. However, the gradient histogram prior used for texture enhancement sometimes generates false textures in the smooth areas. In order to address these problems, the authors propose a robust algorithm combining gradient histogram with sparse representation to obtain good estimates of the sparse coding coefficients of the latent image and realising image denoising while preserving the texture. The proposed model is solved by having a balance between over-enhancement and over-smoothing of the texture in order to preserve the natural texture appearance. Experimental results demonstrate the efficiency and effectiveness of the proposed method.

Stochastic 3d motion compensation of coronary arteries from monoplane angiograms

Image-based navigation during percutaneous coronary interventions is highly challenging since it involves estimating the 3D motion of a complex topology using 2D angiographic views. A static coronary tree segmented in a pre-operative CT-scan can be overlaid on top of the angiographic frames to outline the coronary vessels, but this overlay does not account for coronary motion, which has to be mentally compensated by the cardiologist. In this paper, we propose a new approach to the motion estimation problem, where the temporal evolution of the coronary deformation over the cardiac cycle is modeled as a stochastic process. The sequence of angiographic frames is interpreted as a probabilistic evidence of the succession of unknown deformation states, which can be optimized using particle filtering. Iterative and non-rigid registration is performed in a projective manner, and relies on a feature-based similarity measure. Experiments show promising results in terms of registration accuracy, learning capability and computation time.