Abstract
目的：根据中医“三年化疫＂的运气理论，建立反向传播算法 (BP) 神经网络的传染病预测模型，探索香港呼吸道传染病 (respiratory infectious disease, RID) 与气象因素之间的相关性。
方法：根据香港卫生署卫生防护中心提供的1997-2019年RID的发病数据与香港天文台提供的1994—2019年的平均气温、相对湿度、降雨量等9个气象因素，利用BP人工神经网络的统计学方法，探索性构建＂RID－运气－气象＂预测模型。
结果：香港地区RID在三之气（5—6月份）高发；利用2年前至当年（共3年）的气象因素建立的RID发病预测模型效果最佳，其中的关键因素是平均水汽压。
结论：香港RID发病与前期（如：1~2年前）的气象因素存在一定的相关性，这在一定程度上证明了“三年化疫＂学说的客观性，可能为香港今后的RID预警发挥作用。
Objective: To establish a respiratory infectious disease (RID) prediction model of BP neural network based on the Yunqi theory of ’pestilence occurring after 3 years’, and to explore the correlation between RID and meteorological factors in Hong Kong.
Methods: According to the RID data from Hong Kong Centre For Health during 1997~2019 and meteorological factors (average air temperature, relative humidity, rainfall, 9 meteorological factors) from Hong Kong Observatory during 1994~2019, a prediction model was established through BP artificial neural network thus to explore the correlation between meteorological factors and RID in Hong Kong.
Results: The incidence of RID was the highest in the 3rd qi (May~June). The prediction of RID was successfully established based on the meteorological factors from 2 years ago to the current year (three years in total). The important factor was average air pressure.
Conclusion: There is a certain correlation between the incidence of RID in Hong Kong and the previous meteorological factors, which proves the objectivity of ’pestilence occurring after 3 years’. This concept will help preventing and predicting RID in Hong Kong.
方法：根据香港卫生署卫生防护中心提供的1997-2019年RID的发病数据与香港天文台提供的1994—2019年的平均气温、相对湿度、降雨量等9个气象因素，利用BP人工神经网络的统计学方法，探索性构建＂RID－运气－气象＂预测模型。
结果：香港地区RID在三之气（5—6月份）高发；利用2年前至当年（共3年）的气象因素建立的RID发病预测模型效果最佳，其中的关键因素是平均水汽压。
结论：香港RID发病与前期（如：1~2年前）的气象因素存在一定的相关性，这在一定程度上证明了“三年化疫＂学说的客观性，可能为香港今后的RID预警发挥作用。
Objective: To establish a respiratory infectious disease (RID) prediction model of BP neural network based on the Yunqi theory of ’pestilence occurring after 3 years’, and to explore the correlation between RID and meteorological factors in Hong Kong.
Methods: According to the RID data from Hong Kong Centre For Health during 1997~2019 and meteorological factors (average air temperature, relative humidity, rainfall, 9 meteorological factors) from Hong Kong Observatory during 1994~2019, a prediction model was established through BP artificial neural network thus to explore the correlation between meteorological factors and RID in Hong Kong.
Results: The incidence of RID was the highest in the 3rd qi (May~June). The prediction of RID was successfully established based on the meteorological factors from 2 years ago to the current year (three years in total). The important factor was average air pressure.
Conclusion: There is a certain correlation between the incidence of RID in Hong Kong and the previous meteorological factors, which proves the objectivity of ’pestilence occurring after 3 years’. This concept will help preventing and predicting RID in Hong Kong.
| Translated title of the contribution | Correlation between Meteorological Factors and Incidence of Respiratory Infections in Hong Kong Based on Yunqi Theory |
|---|---|
| Original language | Chinese (Simplified) |
| Pages (from-to) | 1-6 |
| Number of pages | 6 |
| Journal | 中医药学报 |
| Volume | 49 |
| Issue number | 7 |
| DOIs | |
| Publication status | Published - Jul 2021 |
User-Defined Keywords
- 运气理论
- 三年化疫
- 呼吸道传染病
- 香港
- Yunqi theory
- Pestilence occurring after 3 years
- Respiratory infectious diseases
- Hong Kong