395 research outputs found
Automated Characterization of Mobile Health Apps' Features by Extracting Information From the Web: An Exploratory Study
No full textPurpose: The aim of this study was to test the viability of a novel method for automated characterization of mobile health apps. Method: In this exploratory study, we developed the basic modules of an automated method, based on text analytics, able to characterize the apps' medical specialties by extracting information from the web. We analyzed apps in the Medical and Health & Fitness categories on the U.S. iTunes store. Results: We automatically crawled 42,007 Medical and 79,557 Health & Fitness apps' webpages. After removing duplicates and non-English apps, the database included 80,490 apps. We tested the accuracy of the automated method on a subset of 400 apps. We observed 91% accuracy for the identification of apps related to health or medicine, 95% accuracy for sensory systems apps, and an average of 82% accuracy for classification into medical specialties. Conclusions: These preliminary results suggested the viability of automated characterization of apps based on text analytics and highlighted directions for improvement in terms of classification rules and vocabularies, analysis of semantic types, and extraction of key features (promoters, services, and users). The availability of automated tools for app characterization is important as it may support health care professionals in informed, aware selection of health apps to recommend to their patients
Economic factors affecting obesity: an application in Italy
No full textThe World Health Organization has stated that obesity is spreading around the world like a “global epidemic”. In 2004 the percentage of obese people in the Italian population was 9%, but the trend s increasing in recent years. Focusing on this country, the purpose of the paper is to analyze the socio-economic variables affecting obesity by means of a survey conducted in a consumer sample. Our analysis is based on a survey conducted in Italy, and the sample was composed of 999 consumers. We used a binary logit model and the dependent variable is body mass index (BMI), expressed in a dichotomic way (seriously overweight and obese, value 1, and normal weight, value 0). The results show that the condition of the seriously overweight and obese increases with age, especially in people over 65 of age. Also gender is correlated with the pathology: being seriously overweight and obese is far more likely for men than for women. An inverse relation was shown between obesity and education, and between obesity and the level of food knowledge. The results highlight that disadvantaged social categories are more susceptible to the problem of overweight and obesity. A policy implication of the analysis, to limit the spread of obesity, could lie in programs aimed at improving health and food awareness and focused on these minority groups.economics of obesity, BMI and consumer, logit model, Food Consumption/Nutrition/Food Safety, Health Economics and Policy,
Identifying prognostic factors for survival in intensive care unit patients with SIRS or sepsis by machine learning analysis on electronic health records.
No full textBackgroundSystemic inflammatory response syndrome (SIRS) and sepsis are the most common causes of in-hospital death. However, the characteristics associated with the improvement in the patient conditions during the ICU stay were not fully elucidated for each population as well as the possible differences between the two.GoalThe aim of this study is to highlight the differences between the prognostic clinical features for the survival of patients diagnosed with SIRS and those of patients diagnosed with sepsis by using a multi-variable predictive modeling approach with a reduced set of easily available measurements collected at the admission to the intensive care unit (ICU).MethodsData were collected from 1,257 patients (816 non-sepsis SIRS and 441 sepsis) admitted to the ICU. We compared the performance of five machine learning models in predicting patient survival. Matthews correlation coefficient (MCC) was used to evaluate model performances and feature importance, and by applying Monte Carlo stratified Cross-Validation.ResultsExtreme Gradient Boosting (MCC = 0.489) and Logistic Regression (MCC = 0.533) achieved the highest results for SIRS and sepsis cohorts, respectively. In order of importance, APACHE II, mean platelet volume (MPV), eosinophil counts (EoC), and C-reactive protein (CRP) showed higher importance for predicting sepsis patient survival, whereas, SOFA, APACHE II, platelet counts (PLTC), and CRP obtained higher importance in the SIRS cohort.ConclusionBy using complete blood count parameters as predictors of ICU patient survival, machine learning models can accurately predict the survival of SIRS and sepsis ICU patients. Interestingly, feature importance highlights the role of CRP and APACHE II in both SIRS and sepsis populations. In addition, MPV and EoC are shown to be important features for the sepsis population only, whereas SOFA and PLTC have higher importance for SIRS patients
Characterization of Synthetic Health Data Using Rule-Based Artificial Intelligence Models
No full textThe aim of this study is to apply and characterize eXplainable AI (XAI) to assess the quality of synthetic health data generated using a data augmentation algorithm. In this exploratory study, several synthetic datasets are generated using various configurations of a conditional Generative Adversarial Network (GAN) from a set of 156 observations related to adult hearing screening. A rule-based native XAI algorithm, the Logic Learning Machine, is used in combination with conventional utility metrics. The classification performance in different conditions is assessed: models trained and tested on synthetic data, models trained on synthetic data and tested on real data, and models trained on real data and tested on synthetic data. The rules extracted from real and synthetic data are then compared using a rule similarity metric. The results indicate that XAI may be used to assess the quality of synthetic data by (i) the analysis of classification performance and (ii) the analysis of the rules extracted on real and synthetic data (number, covering, structure, cut-off values, and similarity). These results suggest that XAI can be used in an original way to assess synthetic health data and extract knowledge about the mechanisms underlying the generated data
Multivariate Classification of Mild and Moderate Hearing Loss Using a Speech-in-Noise Test for Hearing Screening at a Distance
No full textIn the area of smartphone-based hearing screening, the number of speech-in-noise tests available is growing rapidly. However, the available tests are typically based on a univariate classification approach, for example using the speech recognition threshold (SRT) or the number of correct responses. There is still lack of multivariate approaches to screen for hearing loss (HL). Moreover, all the screening methods developed so far do not assess the degree of HL, despite the potential importance of this information in terms of patient education and clinical follow-up. The aim of this study was to characterize multivariate approaches to identify mild and moderate HL using a recently developed, validated speech-in-noise test for hearing screening at a distance, namely the WHISPER (Widespread Hearing Impairment Screening and PrEvention of Risk) test. The WHISPER test is automated, minimally dependent on the listeners’ native language, it is based on an optimized, efficient adaptive procedure, and it uses a multivariate approach. The results showed that age and SRT were the features with highest performance in identifying mild and moderate HL, respectively. Multivariate classifiers using all the WHISPER features achieved better performance than univariate classifiers, reaching an accuracy equal to 0.82 and 0.87 for mild and moderate HL, respectively. Overall, this study suggested that mild and moderate HL may be discriminated with high accuracy using a set of features extracted from the WHISPER test, laying the ground for the development of future self-administered speech-in-noise tests able to provide specific recommendations based on the degree of HL
Dual-View Single-Shot Multibox Detector at Urban Intersections: Settings and Performance Evaluation
No full textThe explosion of artificial intelligence methods has paved the way for more sophisticated smart mobility solutions. In this work, we present a multi-camera video content analysis (VCA) system that exploits a single-shot multibox detector (SSD) network to detect vehicles, riders, and pedestrians and triggers alerts to drivers of public transportation vehicles approaching the surveilled area. The evaluation of the VCA system will address both detection and alert generation performance by combining visual and quantitative approaches. Starting from a SSD model trained for a single camera, we added a second one, under a different field of view (FOV) to improve the accuracy and reliability of the system. Due to real-time constraints, the complexity of the VCA system must be limited, thus calling for a simple multi-view fusion method. According to the experimental test-bed, the use of two cameras achieves a better balance between precision (68%) and recall (84%) with respect to the use of a single camera (i.e., 62% precision and 86% recall). In addition, a system evaluation in temporal terms is provided, showing that missed alerts (false negatives) and wrong alerts (false positives) are typically transitory events. Therefore, adding spatial and temporal redundancy increases the overall reliability of the VCA system
Computational exposure assessment of electromagnetic fields generated by an RFID system for mother-newborn identity reconfirmation
No full textEnhancement of neural stochastic firing in cochlear implant stimulation by the addition of noise: a computational study of the influence of stimulation settings and spontaneous activity.
No full textPreliminary Evaluation of a Novel Language Independent Speech-in-Noise Test for Adult Hearing Screening
No full textThis article presents a preliminary evaluation of a novel language in- dependent Speech-in-Noise test for adult screening in terms of Speech Reception Threshold (SRT) estimates and prediction of hearing sensitivity.
The test is based on multiple-choice recognition of meaningless Vowel-Consonant-Vowel words and was administered to 26 normal hearing young adults and 58 unscreened adults who also underwent pure-tone audiometry. Receiver operating characteristics were built using the World Health Organization criteria for “slight/mild” and “moderate” hearing loss as gold standards and SRTs as test outcome. Both curves showed very good test performance in predicting success/failure in pure-tone audiometry (area under the curve: 0.79 for “slight/mild” and 0.83 for “moderate” hearing loss). A complete generalized linear model including SRT, age, and their interaction showed that the SRT and the interaction between SRT and age were significant predictors of pure-tone audiometry out- comes, whereas age alone was not a significant predictor of the degree of hearing loss. Moreover, preliminary results from test-retest data showed that the test was reliable in repeated measures (Spearman’s rank-order correlation coefficient = 0.72; Cohen’s kappa = 0.83 for “slight/mild” and 0.64 for “moderate” hearing loss). Further research is needed to fully assess test performance in a larger sample of participants, also including subjects with higher degrees of hearing loss (e.g. “severe” and “profound”)
Automated identification of health apps' medical specialties and promoters from the store webpages
No full textThe aim of this study was to develop automated methods, based on text analytics, for extracting information from the apps' webpages on the app stores and identify relevant apps' features such as the medical specialty and promoter. In this preliminary study, we classified a sample of more than 66000 apps from the US iTunes store into 18 medical specialties and seven types of promoters. Of the â1⁄466000 apps analyzed over 18 specialties, we found that 24.1% were relevant to Nutrition, 23.9% to General Medicine, and 15.7% to Pharmacology, whereas less than 1.5% of apps were relevant to specialties such as Rheumatology, Radiology, Diabetes, Respiratory, Vision, and Sleep Healthcare. The analysis of promoters showed that Manufacturers and Software Houses and Independent Developers promoted 99% of apps combined, whereas promoters in the healthcare and science areas (e.g., Government Services, Healthcare Providers, or Scientific and Educational Organizations) still play a minor role. This study highlighted interesting trends and open opportunities in the field of health apps and suggested that the proposed approach might be a basis for future developments of support tools for informed, aware selection and adoption of health apps by patients and healthcare professionals
- …
