近年来，经济的快速发展、技术的进步等对人们思想观念、行为方式产生着巨大的影响，也使得疾病类型呈现多样化，威胁着人们的健康。目前，恶性肿瘤是全球范围内发病及致死率最高的疾病之一。而宫颈癌作为一种常见的女性恶性肿瘤，在威胁女性健康的同时，也为社会带来了沉重的经济负担。本文从我国女性宫颈癌发病及死亡趋势入手，对其影响因素进行分析并对趋势进行预测，旨在为相关医疗卫生部门制定宫颈癌的防控措施提供数据及理论支撑。

首先，本文在对女性宫颈癌发病及死亡数据收集处理的基础上，对其变化趋势进行了描述性分析，利用Joinpoint回归模型进一步分析趋势变化规律。其次，基于年龄-时期-队列模型，利用内生因子法计算各因素的效应系数值，探讨三因素对宫颈癌发病风险及死亡风险影响的大小。最后，分别构建差分自回归移动平均模型、灰色模型和误逆差传播神经网络模型对标化发病率和标化死亡率进行数据拟合，利用平均绝对误差、均方误差、平均绝对百分比误差三个评价指标对模型拟合效果进行检验。根据分析结果可以得出，BP神经网络模型的预测结果具有较高的精确度。因此选用BP神经网络对2019-2023年女性宫颈癌的标化发病率和标化死亡率进行预测。

研究发现，2003-2018年中国女性宫颈癌的标化发病率和标化死亡率整体呈上升趋势，受年龄因素影响较大，受时期和队列因素的影响较小，2019-2023年标化发病率和标化死亡率将整体呈下降趋势。因此，加强女性宫颈癌三级筛查和人乳头瘤病毒疫苗接种工作，尤其是55-59岁的高发年龄段，制定具有针对性的防控措施是非常必要的。

In recent years, the rapid growth of economy and progress of technology have a great impact on people’s ideas and behavior, which makes the types of diseases diversified and increasingly threatens people’s health. At the moment, malignant tumor is one of the malady with the highest incidence and fatality rates in the world. And Cervical Cancer is a common female malignant tumor that threatens women’s health while bringing economic burden to the society. In the article, we study the trends of incidence and mortality of female cervical cancer in China, analyze influencing factors and predict the trends. Through the analysis and prediction, they provide the supporting of data and theory for the relevant medical departments to formulate cervical cancer prevention and control measures.

First of all, in this article, based on collecting data on the incidence and mortality of female cervical cancer cases in China, we described the changing trends by descriptive statistics analysis, and used the Joinpoint regression model to further analyze the regularity of trends on the base. Secondly, based on the APC model, we used the Intrinsic Estimator to calculate the value of age, period and cohort effects to explore the influences of each factor on the hazard of the incidence and the mortality. Lastly, the ARIMA model, GM model and BP neural network model were constructed to fit the age-adjusted incidence and age-adjusted mortality. And the models fitting effects were tested by applying evaluation indictors of MAE, MSE and MAPE. According to the analysis, it can be concluded that the prediction results of the BP neural network model have high accuracy. Therefore, the BP neural network is used to predict the age-adjusted incidence and age-adjusted mortality of Cervical Cancer in women from 2019 to 2023.

The results show that the age-adjusted incidence and the age-adjusted mortality of female cervical cancer cases increase as a whole between 2003 and 2018, which are more affected by age, and less affected by period and cohort factors. The trends are predicted that they will decline from 2019 to 2023. Therefore, it is important that cervical cancer screening and HPV vaccination for women should be strengthened, especially the peak age of 55-59, and targeted prevention and control measures should be formulated.