English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Background It was shown that different individual weather conditions are associated with the incidence of acute coronary syndromes (ACS). Despite this, the prediction of the number of ACS depending on the weather conditions in the given place and time is not effective to date. Purpose We sought to investigate whether the artificial intelligence system might be useful in prediction of the prevalence of ACS based on weather conditions. Methods In this study, data of 159307 consecutive patients obtained from National Health Service registry, hospitalized due to ACS in Lesser Poland Province (province area of 15008 km2, population of 3.4 M in 2014) between 2008 and 2018 have been compared with meteorological conditions collected in the Institute of Meteorology and Water Management from five weather stations scattered across Lesser Poland Province. Because of small sample size in three of them, only data from two stations (Krakow-Balice, n=75565, Tarnow, n=30079) were used for further analysis. In four separate seasons, the number of ACS events in each day was compared with meteorological conditions on a given day and six days before. We analysed weather conditions such as: wind 10 metres above ground (W_10), temperature (T), dew point temperature (T_dp), relative humidity 2 metres above the ground (Hum_2), atmospheric pressure reduced to mean sea level (Pres), atmospheric precipitation (Prec), and 3 hours atmospheric pressure changes (Pres_3h). For all parameters extreme (maximum – max, minimum – min) values and ranges of these parameters in each day were analysed. All data were used in a system based on machine learning (Random Forest), which allowed to create a model that predicted the incidence of ACS and to determine importance of each inputted weather parameter in this prediction. Results All weather parameters were divided into machine learning data (70%) and test data (30%) to verify functioning of the model. The correlation between real number of ACS and predicted number of ACS for two meteorological stations for spring ranges from 0.69 to 0.71 with confidence intervals (CI) of 0.63–0.77, for summer the correlation was 0.66–0.75 with CI of 0.59–79, for autumn 0.69–0.74 with CI of 0.63–0.79 and for winter 0.69–0.72 with CI 0.63–0.77 (P&amp;lt;0.0001 for each prediction, example of prediction in the Figure 1A). Among all analysed meteorological parameters the most important in the machine learning were range of relative humidity, range of dew point temperature and maximal relative humidity (Figure 1B). Conclusions Artificial intelligence system seems to be useful in predicting the prevalence of ACS with model based on weather conditions. Figure 1. ACS prediction for summer in Krakow Funding Acknowledgement Type of funding source: None

The artificial intelligence system in prediction of the prevalence of acute coronary syndromes based on weather conditions