Université de Sherbrooke

About: Université de Sherbrooke is a education organization based out in Sherbrooke, Quebec, Canada. It is known for research contribution in the topics: Population & Receptor. The organization has 14922 authors who have published 28783 publications receiving 792511 citations. The organization is also known as: Universite de Sherbrooke & Sherbrooke University.

MADE: Masked Autoencoder for Distribution Estimation

Abstract: There has been a lot of recent interest in designing neural network models to estimate a distribution from a set of examples. We introduce a simple modification for autoencoder neural networks that yields powerful generative models. Our method masks the autoencoder's parameters to respect autoregressive constraints: each input is reconstructed only from previous inputs in a given ordering. Constrained this way, the autoencoder outputs can be interpreted as a set of conditional probabilities, and their product, the full joint probability. We can also train a single network that can decompose the joint probability in multiple different orderings. Our simple framework can be applied to multiple architectures, including deep ones. Vectorized implementations, such as on GPUs, are simple and fast. Experiments demonstrate that this approach is competitive with state-of-the-art tractable distribution estimators. At test time, the method is significantly faster and scales better than other autoregressive estimators.

Current status of phthalocyanines in the photodynamic therapy of cancer

Abstract: Photodynamic therapy is a binary treatment now accepted in clinic for various malignancies in several countries around the world. Phthalocyanine molecules are second-generation photosensitizers with enhanced photophysical and photochemical properties over those of porphyrins. They have been shown to be phototoxic against a number of cell types and tumor models. A great deal of research has been devoted to the elucidation of their mechanism of action and mode of cell death. The present paper reviews phthalocyanine pre-clinical anti-cancer research with emphasis on phthalocyanine induced apoptosis using a silicon phthalocyanine, Pc4. A brief summary of the latest clinical results using phthalocyanines is presented.

Relationship between multimorbidity and health-related quality of life of patients in primary care.

Abstract: Previous studies about the association of multimorbidity and the health-related quality of life (HRQOL) in primary-care patients are limited because of their reliance on simple counts of diseases from a limited list of diseases and their failure to assess the severity of disease. We evaluated the association while taking into account the severity of the medical conditions based on the Cumulative Illness Rating Scale (CIRS) score, and controlling for potential confounders (age, sex, household income, education, self-perception of economic status, number of people living in the same dwelling, and perceived social support). We randomly selected 238 patients to construct quintiles of increasing multimorbidity (CIRS). Patients completed the 36-item Medical Outcomes study questionnaire (SF-36) to evaluate their HRQOL. Applying bivariate and multivariate linear regression analyses, we used the CIRS as either a continuous or a categorical (quintiles) variable. Use of the CIRS revealed a stronger association of HRQOL with multimorbidity than using a simple count of chronic conditions. Physical more than mental health deteriorated with increasing multimorbidity. Perceived social support and self-perception of economic status were significantly related to all scales of the SF-36 (p < 0.05). Increased multimorbidity adversely affected HRQOL in primary-care adult patients, even when confounding variables were controlled for.

Non-local Deep Features for Salient Object Detection

Abstract: Saliency detection aims to highlight the most relevant objects in an image. Methods using conventional models struggle whenever salient objects are pictured on top of a cluttered background while deep neural nets suffer from excess complexity and slow evaluation speeds. In this paper, we propose a simplified convolutional neural network which combines local and global information through a multi-resolution 4×5 grid structure. Instead of enforcing spacial coherence with a CRF or superpixels as is usually the case, we implemented a loss function inspired by the Mumford-Shah functional which penalizes errors on the boundary. We trained our model on the MSRA-B dataset, and tested it on six different saliency benchmark datasets. Results show that our method is on par with the state-of-the-art while reducing computation time by a factor of 18 to 100 times, enabling near real-time, high performance saliency detection.

Abiotic drivers and plant traits explain landscape-scale patterns in soil microbial communities

Abstract: The controls on aboveground community composition and diversity have been extensively studied, but our understanding of the drivers of belowground microbial communities is relatively lacking, despite their importance for ecosystem functioning. In this study, we fitted statistical models to explain landscape-scale variation in soil microbial community composition using data from 180 sites covering a broad range of grassland types, soil and climatic conditions in England. We found that variation in soil microbial communities was explained by abiotic factors like climate, pH and soil properties. Biotic factors, namely community-weighted means (CWM) of plant functional traits, also explained variation in soil microbial communities. In particular, more bacterial-dominated microbial communities were associated with exploitative plant traits versus fungal-dominated communities with resource-conservative traits, showing that plant functional traits and soil microbial communities are closely related at the landscape scale.