基于Adaboost的阻塞性睡眠呼吸暂停诊断预测模型研究

Research on Diagnostic Prediction Model of Obstructive Sleep Apnea Based on Adaboost

摘要: 阻塞性睡眠呼吸暂停是一种常见的睡眠呼吸疾病，该疾病由上气道部分或完全塌陷引起，会导致多种疾病和并发症的发生，诊断阻塞性睡眠呼吸暂停的金标准为多导睡眠监测，但该种检测方式也存在一些不便之处，如检查费用昂贵、设备舒适度有待提升、使用环境固定、技师培训复杂等，并且多导睡眠监测难以对大多数住院患者做到高效且便捷的筛查。因此，随着近年来人工智能技术的蓬勃发展，本研究提出了基于Adaboost算法建立机器学习模型，用于预测人群患阻塞性睡眠呼吸暂停的情况。实验结果表明，该模型在所有类别上的平均预测准确率为91%。显示出了较高的分类效果，为该领域疾病检查提供了潜在应用价值。

Abstract: Obstructive sleep apnea is a common sleep breathing disease caused by partial or complete collapse of the upper airway, which can lead to a variety of diseases and complications. The gold standard for diagnosing obstructive sleep apnea is polysomnography. However, this detection method also has some inconveniences, such as expensive inspection costs, equipment comfort needs to be improved, the use of a fixed environment, and the training of technicians is complex, and polysomnography monitoring is difficult to achieve efficient and convenient screening for most hospitalized patients. Therefore, with the vigorous development of artificial intelligence technology in recent years, this study proposed to establish a machine learning model based on Adaboost algorithm to predict the situation of people suffering from obstructive sleep apnea. The experimental results show that the model has an average prediction accuracy of 91% across all categories. It shows a high classification effect and provides potential application value for disease examination in this field.