Search

Quantifying the distribution and morphology of tau protein structures in brain tissues is key to diagnosing Alzheimer's Disease (AD) and its subtypes. Recently, deep learning (DL) models such as UNet have been successfully used for automatic segmentation of histopathological whole slide images (WSI) of biological tissues. In this study, we propose a DL-based methodology for semantic segmentation of tau lesions (i.e., neuritic plaques) in WSI of postmortem patients with AD. The state of the art in semantic segmentation of neuritic plaques in human WSI is very limited. Our study proposes a baseline able to generate a significant advantage for morphological analysis of these tauopathies for further stratification of AD patients. Essential discussions concerning biomarkers (ALZ50 versus AT8 tau antibodies), the imaging modality (different slide scanner resolutions), and the challenge of weak annotations are addressed within this seminal study. The analysis of the impact of context in plaque segmentation is important to understand the role of the micro-environment for reliable tau protein segmentation. In addition, by integrating visual interpretability, we are able to explain how the network focuses on a region of interest (ROI), giving additional insights to pathologists. Finally, the release of a new expert-annotated database and the code (\url{https://github.com/aramis-lab/miccai2022-stratifiad.git}) will be helpful for the scientific community to accelerate the development of new pipelines for human WSI processing in AD.

Cryptographic algorithm implementations are vulnerable to Cold Boot attacks, which consist in exploiting the persistence of RAM cells across reboots or power down cycles to read the memory contents and recover precious sensitive data. The principal defensive weapon against Cold Boot attacks is memory encryption. In this work we propose MemShield, a memory encryption framework for user space applications that exploits a GPU to safely store the master key and perform the encryption/decryption operations. We developed a prototype that is completely transparent to existing applications and does not require changes to the OS kernel. We discuss the design, the related works, the implementation, the security analysis, and the performances of MemShield., Comment: 14 pages, 2 figures. In proceedings of the 18th International Conference on Applied Cryptography and Network Security, ACNS 2020, October 19-22 2020, Rome, Italy

Over the past several decades, naturally occurring and man-made mass casualty incidents (MCI) have increased in frequency and number, worldwide. To test the impact of such event on medical resources, simulations can provide a safe, controlled setting while replicating the chaotic environment typical of an actual disaster. A standardised method to collect and analyse data from mass casualty exercises is needed, in order to assess preparedness and performance of the healthcare staff involved. We report on the use of wearable proximity sensors to measure proximity events during a MCI simulation. We investigated the interactions between medical staff and patients, to evaluate the time dedicated by the medical staff with respect to the severity of the injury of the victims depending on the roles. We estimated the presence of the patients in the different spaces of the field hospital, in order to study the patients' flow. Data were obtained and collected through the deployment of wearable proximity sensors during a mass casualty incident functional exercise. The scenario included two areas: the accident site and the Advanced Medical Post (AMP), and the exercise lasted 3 hours. A total of 238 participants simulating medical staff and victims were involved. Each participant wore a proximity sensor and 30 fixed devices were placed in the field hospital. The contact networks show a heterogeneous distribution of the cumulative time spent in proximity by participants. We obtained contact matrices based on cumulative time spent in proximity between victims and the rescuers. Our results showed that the time spent in proximity by the healthcare teams with the victims is related to the severity of the patient's injury. The analysis of patients' flow showed that the presence of patients in the rooms of the hospital is consistent with triage code and diagnosis, and no obvious bottlenecks were found.

Past UV and optical observations of stars hosting hot Jupiters have shown that some of these stars present an anomalously low chromospheric activity, significantly below the basal level. For WASP-13, observations have shown that the apparent lack of activity is possibly caused by absorption from the intervening ISM. Inspired by this result, we study the effect of ISM absorption on activity measurements (S and logR'$_{\rm HK}$ indices) for main-sequence late-type stars. To this end, we employ synthetic stellar photospheric spectra combined with varying amounts of chromospheric emission and ISM absorption. We present the effect of ISM absorption on activity measurements by varying several instrumental, stellar, and ISM parameters. We find that for relative velocities between the stellar and ISM lines smaller than 30-40 km/s and for ISM CaII column densities logN$_{\rm CaII}$>12, the ISM absorption has a significant influence on activity measurements. Direct measurements and three dimensional maps of the Galactic ISM absorption indicate that an ISM CaII column density of logN$_{\rm CaII}$=12 is typically reached by a distance of about 100 pc along most sight lines. In particular, for a Sun-like star lying at a distance greater than 100 pc, we expect a depression (bias) in the logR'$_{\rm HK}$ value larger than 0.05-0.1 dex, about the same size as the typical measurement and calibration uncertainties on this parameter. This work shows that the bias introduced by ISM absorption must always be considered when measuring activity for stars lying beyond 100 pc. We also consider the effect of multiple ISM absorption components. We discuss the relevance of this result for exoplanet studies., Comment: Accepted for publication in A&A

The chromospheric activity index logR'HK of stars hosting transiting hot Jupiters appears to be correlated with the planets' surface gravity. One of the possible explanations is based on the presence of condensations of planetary evaporated material located in a circumstellar cloud that absorbs the CaII H&K and MgII h&k resonance line emission flux, used to measure chromospheric activity. A larger column density in the condensations, or equivalently a stronger absorption in the chromospheric lines, is obtained when the evaporation rate of the planet is larger, which occurs for a lower gravity of the planet. We analyze here a sample of stars hosting transiting hot Jupiters tuned in order to minimize systematic effects (e.g., interstellar medium absorption). Using a mixture model, we find that the data are best fit by a two-linear-regression model. We interpret this result in terms of the Vaughan-Preston gap. We use a Monte Carlo approach to best take into account the uncertainties, finding that the two intercepts fit the observed peaks of the distribution of logR'HK for main-sequence solar-like stars. We also find that the intercepts are correlated with the slopes, as predicted by the model based on the condensations of planetary evaporated material. Our findings bring further support to this model, although we cannot firmly exclude different explanations. A precise determination of the slopes of the two linear components would allow one to estimate the average effective stellar flux powering planetary evaporation, which can then be used for theoretical population and evolution studies of close-in planets., Comment: 23 pages, 4 figures, 1 table, accepted for publication in ApJL

Mixtures of Gaussian factors are powerful tools for modeling an unobserved heterogeneous population, offering - at the same time - dimension reduction and model-based clustering. Unfortunately, the high prevalence of spurious solutions and the disturbing effects of outlying observations, along maximum likelihood estimation, open serious issues. In this paper we consider restrictions for the component covariances, to avoid spurious solutions, and trimming, to provide robustness against violations of normality assumptions of the underlying latent factors. A detailed AECM algorithm for this new approach is presented. Simulation results and an application to the AIS dataset show the aim and effectiveness of the proposed methodology.

A robust estimator for a wide family of mixtures of linear regression is presented. Robustness is based on the joint adoption of the Cluster Weighted Model and of an estimator based on trimming and restrictions. The selected model provides the conditional distribution of the response for each group, as in mixtures of regression, and further supplies local distributions for the explanatory variables. A novel version of the restrictions has been devised, under this model, for separately controlling the two sources of variability identified in it. This proposal avoids singularities in the log-likelihood, caused by approximate local collinearity in the explanatory variables or local exact fit in regressions, and reduces the occurrence of spurious local maximizers. In a natural way, due to the interaction between the model and the estimator, the procedure is able to resist the harmful influence of bad leverage points along the estimation of the mixture of regressions, which is still an open issue in the literature. The given methodology defines a well-posed statistical problem, whose estimator exists and is consistent to the corresponding solution of the population optimum, under widely general conditions. A feasible EM algorithm has also been provided to obtain the corresponding estimation. Many simulated examples and two real datasets have been chosen to show the ability of the procedure, on the one hand, to detect anomalous data, and, on the other hand, to identify the real cluster regressions without the influence of contamination.

A family of parsimonious Gaussian cluster-weighted models is presented. This family concerns a multivariate extension to cluster-weighted modelling that can account for correlations between multivariate responses. Parsimony is attained by constraining parts of an eigen-decomposition imposed on the component covariance matrices. A sufficient condition for identifiability is provided and an expectation-maximization algorithm is presented for parameter estimation. Model performance is investigated on both synthetic and classical real data sets and compared with some popular approaches. Finally, accounting for linear dependencies in the presence of a linear regression structure is shown to offer better performance, vis-\`{a}-vis clustering, over existing methodologies.

The cluster-weighted model (CWM) is a mixture model with random covariates which allows for flexible clustering and density estimation of a random vector composed by a response variable and by a set of covariates. In this class of models, the generalized linear exponential CWM is here introduced especially for modeling bivariate data of mixed-type. Its natural counterpart, in the family of latent class models, is also defined. Maximum likelihood parameter estimates are derived using the EM algorithm and model selection is carried out using the Bayesian information criterion (BIC). Artificial and real data are finally considered to exemplify and appreciate the proposed model., Comment: the paper has been withdrawn by the authors because they are working on a quite extended version of the paper

Mixtures of factor analyzers are becoming more and more popular in the area of model based clustering of high-dimensional data. According to the likelihood approach in data modeling, it is well known that the unconstrained log-likelihood function may present spurious maxima and singularities and this is due to specific patterns of the estimated covariance structure, when their determinant approaches 0. To reduce such drawbacks, in this paper we introduce a procedure for the parameter estimation of mixtures of factor analyzers, which maximizes the likelihood function in a constrained parameter space. We then analyze and measure its performance, compared to the usual non-constrained approach, via some simulations and applications to real data sets., Comment: 21 pages, 7 figures, 8 tables

We consider multivariate skew-t distributions for modeling composition data of high energy cosmic rays. The model has been validated with simulated data for different primary nuclei and hadronic models focusing on the depth of maximum Xmax and number of muons N{\mu} observables. Further, we consider mixtures of multivariate skew-t distributions for cosmic ray mass composition determination and event-by-event classification. With respect to other approaches in the field, it is based on analytical calculations and allows to incorporate different sets of constraints provided by the present hadronic models. We present some applications to simulated data sets generated with different nuclear abundances assumptions. As it does not fully rely on the hadronic model predictions, the method is particularly suited to the current experimental scenario in which evidences of discrepancies of the measured data with respect to the models have been reported for some shower observables, such as the number of muons at ground level., Comment: 17 pages

Cluster-Weighted Modeling (CWM) is a flexible mixture approach for modeling the joint probability of data coming from a heterogeneous population as a weighted sum of the products of marginal distributions and conditional distributions. In this paper, we introduce a wide family of Cluster Weighted models in which the conditional distributions are assumed to belong to the exponential family with canonical links which will be referred to as Generalized Linear Gaussian Cluster Weighted Models. Moreover, we show that, in a suitable sense, mixtures of generalized linear models can be considered as nested in Generalized Linear Gaussian Cluster Weighted Models. The proposal is illustrated through many numerical studies based on both simulated and real data sets.

In model-based clustering and classification, the cluster-weighted model constitutes a convenient approach when the random vector of interest constitutes a response variable Y and a set p of explanatory variables X. However, its applicability may be limited when p is high. To overcome this problem, this paper assumes a latent factor structure for X in each mixture component. This leads to the cluster-weighted factor analyzers (CWFA) model. By imposing constraints on the variance of Y and the covariance matrix of X, a novel family of sixteen CWFA models is introduced for model-based clustering and classification. The alternating expectation-conditional maximization algorithm, for maximum likelihood estimation of the parameters of all the models in the family, is described; to initialize the algorithm, a 5-step hierarchical procedure is proposed, which uses the nested structures of the models within the family and thus guarantees the natural ranking among the sixteen likelihoods. Artificial and real data show that these models have very good clustering and classification performance and that the algorithm is able to recover the parameters very well., Comment: 36 pages, 6 figures