Comparative study of convolutional neural network architectures for gastrointestinal lesions classification.

The gastrointestinal (GI) tract can be affected by different diseases or lesions such as esophagitis, ulcers, hemorrhoids, and polyps, among others. Some of them can be precursors of cancer such as polyps. Endoscopy is the standard procedure for the detection of these lesions. The main drawback of this procedure is that the diagnosis depends on the expertise of the doctor. This means that some important findings may be missed. In recent years, this problem has been addressed by deep learning (DL) techniques. Endoscopic studies use digital images. The most widely used DL technique for image processing is the convolutional neural network (CNN) due to its high accuracy for modeling complex phenomena. There are different CNNs that are characterized by their architecture. In this article, four architectures are compared: AlexNet, DenseNet-201, Inception-v3, and ResNet-101. To determine which architecture best classifies GI tract lesions, a set of metrics; accuracy, precision, sensitivity, specificity, F1-score, and area under the curve (AUC) were used. These architectures were trained and tested on the HyperKvasir dataset. From this dataset, a total of 6,792 images corresponding to 10 findings were used. A transfer learning approach and a data augmentation technique were applied. The best performing architecture was DenseNet-201, whose results were: 97.11% of accuracy, 96.3% sensitivity, 99.67% specificity, and 95% AUC.

Univariate Analysis of Short-Chain Fatty Acids Related to Sudden Infant Death Syndrome.

Sudden infant death syndrome (SIDS) is defined as the death of a child under one year of age, during sleep, without apparent cause, after exhaustive investigation, so it is a diagnosis of exclusion. SIDS is the principal cause of death in industrialized countries. Inborn errors of metabolism (IEM) have been related to SIDS. These errors are a group of conditions characterized by the accumulation of toxic substances usually produced by an enzyme defect and there are thousands of them and included are the disorders of the ß-oxidation cycle, similarly to what can affect the metabolism of different types of fatty acid chain (within these, short chain fatty acids (SCFAs)). In this work, an analysis of postmortem SCFAs profiles of children who died due to SIDS is proposed. Initially, a set of features containing SCFAs information, obtained from the NIH Common Fund's National Metabolomics Data Repository (NMDR) is submitted to an univariate analysis, developing a model based on the relationship between each feature and the binary output (death due to SIDS or not), obtaining 11 univariate models. Then, each model is validated, calculating their receiver operating characteristic curve (ROC curve) and area under the ROC curve (AUC) value. For those features whose models presented an AUC value higher than 0.650, a new multivariate model is constructed, in order to validate its behavior in comparison to the univariate models. In addition, a comparison between this multivariate model and a model developed based on the whole set of features is finally performed. From the results, it can be observed that each SCFA which comprises of the SFCAs profile, has a relationship with SIDS and could help in risk identification.

Demographic and Comorbidities Data Description of Population in Mexico with SARS-CoV-2 Infected Patients(COVID19): An Online Tool Analysis.

The Word Health Organization (WHO) declared in March 2020 that we are facing a pandemic designated as COVID-19, which is the acronym of coronavirus disease 2019, caused by a new virus know as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In Mexico, the first cases of COVID-19, was reported by the Secretary of Health on 28 February 2020. More than sixteen thousand cases and more than fifteen thousand deaths have been reported in Mexico, and it continues to rise; therefore, this article proposes two online visualization tools (a web platform) that allow the analysis of demographic data and comorbidities of the Mexican population. The objective of these tools is to provide graphic information, fast and updated, based on dataset obtained directly from National Governments Health Secretary (Secretaría de Salud, SSA) which is daily refreshed with the information related to SARS-CoV-2. To allow a dynamical update and friendly interface, and approach with R-project, a well-known Open Source language and environment for statistical computing and Shiny package, were implemented. The dataset is loaded automatically from the latest version released by the federal government of Mexico. Users can choose to study particular groups determined by gender, entity, type of result (positive, negative, pending outcome) and comorbidity. The image results are plots that can be instantly interpreted and supported by the text summary. This tool, in addition to being a consultation for the general public, is useful in Public Health to facilitate the visualization of the data, allowing its timely interpretation due to the changing nature of COVID-19, it can even be used for decision-making by leaders, for the benefit of the health of the community.

A Case⁻Control Study of Socio-Economic and Nutritional Characteristics as Determinants of Dental Caries in Different Age Groups, Considered as Public Health Problem: Data from NHANES 2013⁻2014.

One of the principal conditions that affects oral health worldwide is dental caries, occurring in about 90% of the global population. This pathology has been considered a challenge because of its high prevalence, besides being a chronic but preventable disease which can be caused by a series of different demographic, dietary" among others. Based on this problem, in this research a demographic and dietary features analysis is performed for the classification of subjects according to their oral health status based on caries, according to the age group where the population belongs, using as feature selector a technique based on fast backward selection (FBS) approach for the development of three predictive models, one for each age range (group 1: 10⁻19; group 2: 20⁻59; group 3: 60 or more years old). As validation, a net reclassification improvement (NRI), AUC, ROC, and OR values are used to evaluate their classification accuracy. We analyzed 189 demographic and dietary features from National Health and Nutrition Examination Survey (NHANES) 2013⁻2014. Each model obtained statistically significant results for most features and narrow OR confidence intervals. Age group 2 obtained a mean NRI = -0.080 and AUC = 0.933; age group 3 obtained a mean NRI = -0.024 and AUC = 0.787; and age group 4 obtained a mean NRI = -0.129 and AUC = 0.735. Based on these results, it is concluded that these specific demographic and dietary features are significant determinants for estimating the oral health status in patients based on their likelihood of developing caries, and the age group could imply different risk factors for subjects.

Speed Bump Detection Using Accelerometric Features: A Genetic Algorithm Approach.

Among the current challenges of the Smart City, traffic management and maintenance are of utmost importance. Road surface monitoring is currently performed by humans, but the road surface condition is one of the main indicators of road quality, and it may drastically affect fuel consumption and the safety of both drivers and pedestrians. Abnormalities in the road, such as manholes and potholes, can cause accidents when not identified by the drivers. Furthermore, human-induced abnormalities, such as speed bumps, could also cause accidents. In addition, while said obstacles ought to be signalized according to specific road regulation, they are not always correctly labeled. Therefore, we developed a novel method for the detection of road abnormalities (i.e., speed bumps). This method makes use of a gyro, an accelerometer, and a GPS sensor mounted in a car. After having the vehicle cruise through several streets, data is retrieved from the sensors. Then, using a cross-validation strategy, a genetic algorithm is used to find a logistic model that accurately detects road abnormalities. The proposed model had an accuracy of 0.9714 in a blind evaluation, with a false positive rate smaller than 0.018, and an area under the receiver operating characteristic curve of 0.9784. This methodology has the potential to detect speed bumps in quasi real-time conditions, and can be used to construct a real-time surface monitoring system.