高级检索

基于随机森林算法的心电图引导手臂输液港静脉导管头端精准定位的影响因素研究

A study on influence factors of accurate positioning of venous catheter tip in arm infusion port guided by electrocardiogram based on random forest algorithm

    摘要
  • 摘要:
      目的  基于随机森林算法探讨心电图技术引导手臂输液港静脉导管头端精准定位的影响因素。
      方法  选取338例植入输液港的女性乳腺癌化疗患者作为研究对象, 应用心电图引导静脉导管头端定位, 并采集患者的相关信息。采用随机森林方法探讨心电图引导手臂输液港静脉导管头端精准定位的影响因素, 应用分类树拟合导管头端定位精准性模型。
      结果  身高、体质量和年龄与导管头端定位精准性相关。所构建模型的准确率(0.84)、灵敏度(0.98)、阳性预测值(0.85)尚可, 特异度(0.19)、阴性预测值(0.54)、曲线下面积(AUC)(0.67)较低。
      结论  临床人员应着重关注低身高(< 155 cm)、高体质量(>52 kg)、年龄49~63岁的乳腺癌患者, 因为此类患者易发生静脉导管头端定位不精准的情况。

     

    Abstract:
      Objective  To explore influence factors of accurate positioning of venous catheter tip in arm infusion port guided by electrocardiogram based on random forest algorithm.
      Methods  A total of 338 female breast cancer patients with chemotherapy implanted in transfusion port were selected as research objects. Electrocardiogram was used to guide the positioning of the venous catheter tip, and relevant information of the patients was collected. Random forest method was used to explore the influencing factors of electrocardiogram guided precise positioning of intravenous catheter tip in arm infusion port, and classification tree was used to fit the precise positioning model of catheter tip.
      Results  Height, weight and age were significantly associated with the accuracy of catheter head positioning. The accuracy (0.84), sensitivity (0.98), positive predictive value (0.85) of the constructed model were acceptable, and the specificity (0.19), negative predictive value (0.54) and area under the Curve (AUC)(0.67) were low.
      Conclusion  Clinical staffshould pay attention to breast cancer patients with low height (< 155 cm), high body mass (>52 kg) and age of 49 to 63 years, because such patients are prone to occur inaccurate positioning of the head of venous catheter.

     

    • HTML全文
    • 参考文献(20)
    • 相关文章
    • 施引文献
  • 资源附件(0)

/

返回文章
返回