JCSE, vol. 11, no. 4, pp.142-151, December, 2017

DOI: http://dx.doi.org/10.5626/JCSE.2017.11.4.142

Use of Information Technologies to Explore Correlations between Climatic Factors and Spontaneous Intracerebral Hemorrhage in Different Age Groups

Hsien-Wei Ting, Chien-Lung Chan, Ren-Hao Pan, Robert K. Lai, and Ting-Ying Chien
Department of Neurosurgery, Taipei Hospital, Ministry of Health and Welfare, New Taipei City, Taiwan; Department of Computer Science and Engineering, Yuan Ze University, Taoyuan City, Taiwan Department of Information Management, Yuan Ze University, Taoyuan City, Taiwan; Innovation Center for Big Data and Digital Convergence, Yuan Ze University, Taoyuan City, Taiwan Department of Computer Science and Engineering, Yuan Ze University, Taoyuan City, Taiwan; Department of Information Management, Tunghai University, Taichung City, Taiwan Department of Computer Science and Engineering, Yuan Ze University, Taoyuan City, Taiwan Department of Computer Science and Engineering, Yuan Ze University, Taoyuan City, Taiwan; Innovation Center for Big Data and Digital Convergence, Yuan Ze University, Taoyuan City, Taiwan

Abstract: Spontaneous intracerebral hemorrhage (sICH) has a high mortality rate. Research has demonstrated that sICH occurrence is related to weather conditions; therefore, this study used the decision tree method to explore the impact of climatic risk factors on sICH at different ages. The Taiwan National Health Insurance Research Database (NHIRD) and other open-access data were used in this study. The inclusion criterion was a first-attack sICH. The decision tree algorithm and random forest were implemented in R programming language. We defined a high risk of sICH as more than the average number of cases daily, and the younger, middle-aged and older groups were calculated as having 0.77, 2.26 and 2.60 cases per day, respectively. In total, 22,684 sICH cases were included in this study; 3,102 patients were younger (<44 years, younger group), 9,089 were middle-aged (45-64 years, middle group), and 10,457 were older (>65 years, older group). The risk of sICH in the younger group was not correlated with temperature, wind speed or humidity. The middle group had two decision nodes: a higher risk if the maximum temperature was >19Celsius(probability = 63.7%), and if the maximum temperature was <19Celsius in addition to a wind speed <2.788 (m/s) (probability = 60.9%). The older group had a higher risk if the average temperature was >23.933Celsius(probability = 60.7%). This study demonstrated that the sICH incidence in the younger patients was not significantly correlated with weather factors; that in the middle-aged sICH patients was highly-correlated with the apparent temperature; and that in the older sICH patients was highly-correlated with the mean ambient temperature. "Warm" cold ambient temperatures resulted in a higher risk of sICH, especially in the older patients.

Keyword: Big data; Climatic factors; Decision tree; Random forest; Spontaneous intracerebral hemorrhage

Full Paper:   379 Downloads, 1482 View