Applying Machine Learning to Finger Movements Using Electromyography and Visualization in Opensim.

Electromyographic signals have been used with low-degree-of-freedom prostheses, and recently with multifunctional prostheses. Currently, they are also being used as inputs in the human-computer interface that controls interaction through hand gestures. Although there is a gap between academic publications on the control of an upper-limb prosthesis developed in laboratories and its service in the natural environment, there are attempts to achieve easier control using multiple muscle signals. This work contributes to this, using a database and biomechanical simulation software, both open access, to seek simplicity in the classifiers, anticipating their implementation in microcontrollers and their execution in real time. Fifteen predefined finger movements of the hand were identified using classic classifiers such as Bayes, linear and quadratic discriminant analysis. The idealized movements of the database were modeled with Opensim for visualization. Combinations of two preprocessing methods-the forward sequential selection method and the feature normalization method-were evaluated to increase the efficiency of these classifiers. The statistical methods of cross-validation, analysis of variance (ANOVA) and Duncan were used to validate the results. Furthermore, the classifier with the best recognition result was redesigned into a new feature space using the sparse matrix algorithm to improve it, and to determine which features can be eliminated without degrading the classification. The classifiers yielded promising results-the quadratic discriminant being the best, achieving an average recognition rate for each individual considered of 96.16%, and with 78.36% for the total sample group of the eight subjects, in an independent test dataset. The study ends with the visual analysis under Opensim of the classified movements, in which the usefulness of this simulation tool is appreciated by revealing the muscular participation, which can be useful during the design of a multifunctional prosthesis.

Bayesian Classification Models for Premature Ventricular Contraction Detection on ECG Traces.

According to the American Heart Association, in its latest commission about Ventricular Arrhythmias and Sudden Death 2006, the epidemiology of the ventricular arrhythmias ranges from a series of risk descriptors and clinical markers that go from ventricular premature complexes and nonsustained ventricular tachycardia to sudden cardiac death due to ventricular tachycardia in patients with or without clinical history. The premature ventricular complexes (PVCs) are known to be associated with malignant ventricular arrhythmias and sudden cardiac death (SCD) cases. Detecting this kind of arrhythmia has been crucial in clinical applications. The electrocardiogram (ECG) is a clinical test used to measure the heart electrical activity for inferences and diagnosis. Analyzing large ECG traces from several thousands of beats has brought the necessity to develop mathematical models that can automatically make assumptions about the heart condition. In this work, 80 different features from 108,653 ECG classified beats of the gold-standard MIT-BIH database were extracted in order to classify the Normal, PVC, and other kind of ECG beats. Three well-known Bayesian classification algorithms were trained and tested using these extracted features. Experimental results show that the F1 scores for each class were above 0.95, giving almost the perfect value for the PVC class. This gave us a promising path in the development of automated mechanisms for the detection of PVC complexes.

Genetic Variance Partitioning and Genome-Wide Prediction with Allele Dosage Information in Autotetraploid Potato.

As one of the world's most important food crops, the potato (Solanum tuberosum L.) has spurred innovation in autotetraploid genetics, including in the use of SNP arrays to determine allele dosage at thousands of markers. By combining genotype and pedigree information with phenotype data for economically important traits, the objectives of this study were to (1) partition the genetic variance into additive vs. nonadditive components, and (2) determine the accuracy of genome-wide prediction. Between 2012 and 2017, a training population of 571 clones was evaluated for total yield, specific gravity, and chip fry color. Genomic covariance matrices for additive (G), digenic dominant (D), and additive × additive epistatic (G#G) effects were calculated using 3895 markers, and the numerator relationship matrix (A) was calculated from a 13-generation pedigree. Based on model fit and prediction accuracy, mixed model analysis with G was superior to A for yield and fry color but not specific gravity. The amount of additive genetic variance captured by markers was 20% of the total genetic variance for specific gravity, compared to 45% for yield and fry color. Within the training population, including nonadditive effects improved accuracy and/or bias for all three traits when predicting total genotypic value. When six F1 populations were used for validation, prediction accuracy ranged from 0.06 to 0.63 and was consistently lower (0.13 on average) without allele dosage information. We conclude that genome-wide prediction is feasible in potato and that it will improve selection for breeding value given the substantial amount of nonadditive genetic variance in elite germplasm.

Tracking the Evolution of the Internet of Things Concept Across Different Application Domains.

Both the idea and technology for connecting sensors and actuators to a network to remotely monitor and control physical systems have been known for many years and developed accordingly. However, a little more than a decade ago the concept of the Internet of Things (IoT) was coined and used to integrate such approaches into a common framework. Technology has been constantly evolving and so has the concept of the Internet of Things, incorporating new terminology appropriate to technological advances and different application domains. This paper presents the changes that the IoT has undertaken since its conception and research on how technological advances have shaped it and fostered the arising of derived names suitable to specific domains. A two-step literature review through major publishers and indexing databases was conducted; first by searching for proposals on the Internet of Things concept and analyzing them to find similarities, differences, and technological features that allow us to create a timeline showing its development; in the second step the most mentioned names given to the IoT for specific domains, as well as closely related concepts were identified and briefly analyzed. The study confirms the claim that a consensus on the IoT definition has not yet been reached, as enabling technology keeps evolving and new application domains are being proposed. However, recent changes have been relatively moderated, and its variations on application domains are clearly differentiated, with data and data technologies playing an important role in the IoT landscape.

Glucose Oxidase Biosensor Modeling and Predictors Optimization by Machine Learning Methods.

Biosensors are small analytical devices incorporating a biological recognition element and a physico-chemical transducer to convert a biological signal into an electrical reading. Nowadays, their technological appeal resides in their fast performance, high sensitivity and continuous measuring capabilities; however, a full understanding is still under research. This paper aims to contribute to this growing field of biotechnology, with a focus on Glucose-Oxidase Biosensor (GOB) modeling through statistical learning methods from a regression perspective. We model the amperometric response of a GOB with dependent variables under different conditions, such as temperature, benzoquinone, pH and glucose concentrations, by means of several machine learning algorithms. Since the sensitivity of a GOB response is strongly related to these dependent variables, their interactions should be optimized to maximize the output signal, for which a genetic algorithm and simulated annealing are used. We report a model that shows a good generalization error and is consistent with the optimization.

Gene discovery for facioscapulohumeral muscular dystrophy by machine learning techniques.

Facioscapulohumeral muscular dystrophy (FSHD) is a neuromuscular disorder that shows a preference for the facial, shoulder and upper arm muscles. FSHD affects about one in 20-400,000 people, and no effective therapeutic strategies are known to halt disease progression or reverse muscle weakness or atrophy. Many genes may be incorrectly regulated in affected muscle tissue, but the mechanisms responsible for the progressive muscle weakness remain largely unknown. Although machine learning (ML) has made significant inroads in biomedical disciplines such as cancer research, no reports have yet addressed FSHD analysis using ML techniques. This study explores a specific FSHD data set from a ML perspective. We report results showing a very promising small group of genes that clearly separates FSHD samples from healthy samples. In addition to numerical prediction figures, we show data visualizations and biological evidence illustrating the potential usefulness of these results.

Effective classification and gene expression profiling for the Facioscapulohumeral Muscular Dystrophy.

The Facioscapulohumeral Muscular Dystrophy (FSHD) is an autosomal dominant neuromuscular disorder whose incidence is estimated in about one in 400,000 to one in 20,000. No effective therapeutic strategies are known to halt progression or reverse muscle weakness and atrophy. It is known that the FSHD is caused by modifications located within a D4ZA repeat array in the chromosome 4q, while recent advances have linked these modifications to the DUX4 gene. Unfortunately, the complete mechanisms responsible for the molecular pathogenesis and progressive muscle weakness still remain unknown. Although there are many studies addressing cancer databases from a machine learning perspective, there is no such precedent in the analysis of the FSHD. This study aims to fill this gap by analyzing two specific FSHD databases. A feature selection algorithm is used as the main engine to select genes promoting the highest possible classification capacity. The combination of feature selection and classification aims at obtaining simple models (in terms of very low numbers of genes) capable of good generalization, that may be associated with the disease. We show that the reported method is highly efficient in finding genes to discern between healthy cases (not affected by the FSHD) and FSHD cases, allowing the discovery of very parsimonious models that yield negligible repeated cross-validation error. These models in turn give rise to very simple decision procedures in the form of a decision tree. Current biological evidence regarding these genes shows that they are linked to skeletal muscle processes concerning specific human conditions.

Parsimonious selection of useful genes in microarray gene expression data.

Machine learning methods have of late made significant efforts to solving multidisciplinary problems in the field of cancer classification in microarray gene expression data. These tasks are characterized by a large number of features and a few observations, making the modeling a nontrivial undertaking. In this study, we apply entropic filter methods for gene selection, in combination with several off-the-shelf classifiers. The introduction of bootstrap resampling techniques permits the achievement of more stable performance estimates. Our findings show that the proposed methodology permits a drastic reduction in dimension, offering attractive solutions in terms of both prediction accuracy and number of explanatory genes; a dimensionality reduction technique preserving discrimination capabilities is used for visualization of the selected genes.

Improving attitudes towards children with disabilities in a school context: a cluster randomized intervention study.

AIM: although inclusive education of disabled children is now an accepted practice, it is often challenged by negative peer attitudes. We undertook an interventional study aimed at improving students' attitudes towards their disabled peers. METHOD: the participants were students from the 7th grade of twelve paired schools (1509 students from 62 classes; age 12-13y), randomly allocated to an intervention group (205 males, 285 females) or a control group (132 males, 165 females). The intervention consisted of a mandatory comprehensive educational project on disability. The Chedoke-McMaster Attitudes Towards Children with Handicaps Scale (CATCH) was used to assess children's attitudes before (T0) and after (T1) intervention. The hierarchical structure of the data was taken into account by adjusting standard deviations and using linear multilevel models. RESULTS: seven hundred and eighty-four students had at least one score on the three domains (cognitive, affective, behavioural) of the CATCH at T0 and T1. The final scores were higher than baseline scores (total scores, intervention group: baseline score 25.6 (SD=5.4), final score 26.8 (5.9), p<0.001; CONTROL GROUP: baseline 25.2 (5.4), final 26.0 (5.7), p<0.009) with no significant difference between the intervention and control groups. Individual score changes over time were associated with baseline score (p<0.001 for total and all sub-scores). Lower improvement in attitudes was found in students from schools with special units for their peers with cognitive impairment for total (p=0.013), affective (p<0.001), and behavioural (p=0.001) scores, while higher improvement existed for the cognitive domain (p=0.029). INTERPRETATION: although we found no effect of our intervention, we found an improvement in attitudes in the intervention and control groups that could be a result of the nature of the scales and questionnaires the students had to complete before the intervention.

Determinants of students' attitudes towards peers with disabilities.

AIM: To explore factors associated with students' attitudes towards their peers with disabilities. METHOD: All 7th grade students (aged 12-13y) from 12 schools in the Toulouse area were invited to participate (n=1509). Attitudes were measured using the Chedoke-McMaster Attitudes Towards Children with Handicaps (CATCH) questionnaire (affective, behavioural, cognitive, and total scores). Personal characteristics, including KIDSCREEN quality of life scores, were recorded. Data regarding information about disabilities received from parents and the media and acquaintance with people with disabilities constituted the 'disability knowledge' factors. The characteristics of the schools were obtained from the local education authority. Multivariate multilevel linear regression analyses were conducted to explore the associations between CATCH scores and these three groups of factors. RESULTS: Responses from 1135 students (612 females, 523 males; mean age 12y 8mo SD 7mo; age range 10y 8mo-15y) were studied (75.2% of the students approached). Factors independently associated with more positive attitudes were being a female, having a good quality of life, being friends with a child with disabilities, or having received information about disabilities from parents and the media. Presence in the school of a special class for children with cognitive disabilities was independently associated with more negative attitudes. INTERPRETATION: This cross-sectional study identified different personal and environmental factors upon which interventions aimed at improving students' attitudes towards their peers with disabilities could be based.

Effects of a large-scale industrial disaster on rates of symptoms consistent with posttraumatic stress disorders among schoolchildren in toulouse.

BACKGROUND: Posttraumatic stress disorder (PTSD) has been studied largely among adults and in the context of intentional, collective experiences such as war and terrorism. Far less is known about PTSD among adolescents and resulting from massive industrial accidents. Such an accident in Toulouse, France, 10 days after the World Trade Center disaster, provided an opportunity to examine its effects among adolescents already sensitized by media coverage of the World Trade Center disaster. OBJECTIVES: (1) To assess the presence of symptoms consistent with PTSD (SCW-PTSD) among adolescents in Toulouse after a massive industrial accident, (2) to determine the "excess" of SCW-PTSD among those directly exposed vs those nondirectly exposed, and (3) to examine dosage effects for exposure and the cumulative effect on PTSD of accident-related experiences. DESIGN, SETTING, AND PARTICIPANTS: A survey containing questions on exposure and SCW-PTSD was administered to students aged 11 years, 13 years, 15 years, and 17 years who were enrolled in randomly selected, grade-stratified classrooms from schools for directly exposed students (n = 577) in Toulouse and nondirectly exposed students (n = 900) in the region.Main Outcome Measure The prevalence of SCW-PTSD among directly exposed and nondirectly exposed students. RESULTS: Nine months after the industrial accident, 44.6% of 11- and 13-year-old directly exposed students and 28.5% of 15- and 17-year-old directly exposed students still showed SCW-PTSD, compared with 22.1% of 11- and 13-year-old nondirectly exposed students and 4.4% of 15-year-old nondirectly exposed students. Among 11- and 13-year-olds, the likelihood of having SCW-PTSD was higher for girls who were enrolled in elementary schools, were personally injured, and had severe damage at home, as opposed to boys who were high-school students without severe damage at home or personal injury. Among the 15- and 17-year-olds, being a girl, 17 years old, and personally injured increased the likelihood of having SCW-PTSD, as opposed to 15-year-old boys who were not injured. The effects of injuries were cumulative: students injured personally and with an injured family member were more likely to have SCW-PTSD than those experiencing either personal or family injury but not both. Excess of SCW-PTSD attributable to direct exposure was 50.5% for 11-year-olds, 49.3% for 13-year-olds, and 73.5% for 15-year-olds. CONCLUSIONS: A substantial proportion of Toulouse adolescents still had SCW-PTSD 9 months after the accident. Directly exposed students were far more likely to show SCW-PTSD than those nondirectly exposed, but both groups had SCW-PTSD at rates that were higher than expected. The symptoms were associated with demographic characteristics and direct experiences of trauma. Higher rates applied to students who were personally injured with injured family members and severe damage at home. Students with these characteristics predictive of SCW-PTSD should be given prompt attention to avoid long-lasting effects.