Automated detection and classification of sleep-disordered breathing from conventional polysomnography data.
Automated detection and classification of sleep-disordered breathing from conventional polysomnography data.
复制标题
根据传统多导睡眠图数据自动检测和分类睡眠呼吸障碍。
作者：
Taha,BH;Dempsey,JA;Weber,SM;Badr,MS;Skatrud,JB;Young,TB;Jacques,AJ;Seow,KC
从常规多导睡眠图（PSG）数据中有效地自动检测睡眠呼吸障碍（SDB）是困难的，因为只能间接测量呼吸。我们用来克服这一局限性的方法是将脉搏血氧饱和度纳入呼吸暂停和呼吸不足的定义。在我们的算法中，1）我们开始检测去饱和，一旦下降速率大于每秒0.1%，则氧合血红蛋白饱和度水平下降2%或更大（但低于每秒4%），然后询问是否是呼吸暂停或呼吸不足造成的; 2）如果存在持续至少10秒并且与去饱和事件一致的无呼吸时段，则检测到呼吸暂停，如由总呼吸感应体积描记术（RIP）所指示的;以及3）如果存在呼吸，则呼吸不足被定义为最少三次呼吸，其显示总RIP幅度从紧接的前一次呼吸至少减少20%，随后返回到该“基线”呼吸的至少90%。我们使用包含1，938个SDB事件的10个PSG记录对该算法进行评估，结果显示与经验丰富的多导睡眠仪手动评分的事件一致。人工验证的计算机去饱和度的基础上，检测的灵敏度和特异性百分比分别为73.6%和90.8%的呼吸暂停和84.1%和86.1%的呼吸不足。总体而言，该算法检测到了93.1%的手动检测事件。我们已经设计了一个有效的算法来检测和分类SDB事件，模拟人工评分具有很高的准确性。
Efficient automated detection of sleep-disordered breathing (SDB) from routine polysomnography (PSG) data is made difficult by the availability of only indirect measurements of breathing. The approach we used to overcome this limitation was to incorporate pulse oximetry into the definitions of apnea and hypopnea. In our algorithm, 1) we begin with the detection of desaturation as a fall in oxyhemoglobin saturation level of 2% or greater once a rate of descent greater than 0.1 % per second (but less than 4% per second) has been achieved and then ask if an apnea or hypopnea was responsible; 2) an apnea is detected if there is a period of no breathing, as indicated by sum respiratory inductive plethysmography (RIP), lasting at least 10 seconds and coincident with the desaturation event; and 3) if there is breathing, a hypopnea is defined as a minimum of three breaths showing at least 20% reduction in sum RIP magnitude from the immediately preceding breath followed by a return to at least 90% of that “baseline” breath. Our evaluation of this algorithm using 10 PSG records containing 1,938 SDB events showed strong event-by-event agreement with manual scoring by an experienced polysomnographer. On the basis of manually verified computer desaturations, detection sensitivity and specificity percentages were, respectively, 73.6 and 90.8% for apneas and 84.1 and 86.1% for hypopneas. Overall, 93.1% of the manually detected events were detected by the algorithm. We have designed an efficient algorithm for detecting and classifying SDB events that emulates manual scoring with high accuracy.