基于相对熵的睡眠呼吸暂停低通气综合征的自动诊断

Automated diagnosis of sleep apnea hypopnea syndrome based on relative entropy

目的通过单通道呼吸气流数据,提出一种新的自动诊断睡眠呼吸暂停低通气综合征的方法。方法该方法基于相对熵引入局部极差进行调整,以识别发生呼吸障碍的异常事件。利用75例患者一整晚呼吸气流数据作为训练样本,通过阈值分析,确定区分呼吸暂停事件和低通气事件分别对应的调整后相对熵的临界值。结果用37例患者的数据作为测试样本,验证了基于相对熵方法可以自动诊断病情,而且仅需要20 min左右。结论基于相对熵方法明显缩短了计算时间,对于睡眠疾病的临床诊断有一定的意义。

Objective To present a new approach to the diagnosis of sleep apnea hypopnea syndrome （SAHS） from a single-channel airflow record. Methods Based on relative entropy, the proposed algorithm was adjusted by local range and identified the sleep-disordered breathing automatically. The development of the proposed approach was based on training sets fi＇om the overnight airflow records of 75 patients. Through the threshold analysis, the critical value of the adjusted relative entropy of the apnea event and hypopnea event could be determined, respectively. Results The detection performance of the approach was tested by using records fi＇om another 37 patients. The results suggested that the algorithm may be implemented successfully in automated diagnosis. It took only about 20 minutes for the diagnosis. Conclusions The approach has significantly shortened the diagnosis time in terms of efficiency and has certain significance for the clinical diagnosis in the field of sleep.