AUTOMATED (AI) ANALYSIS OF SLEEP DISORDERED BREATHING

这个子项目是许多研究子项目中的一个 由NIH/NCRR资助的中心赠款提供的资源。子项目和 研究者（PI）可能从另一个NIH来源获得了主要资金， 因此可以在其他CRISP条目中表示。所列机构为 研究中心，而研究中心不一定是研究者所在的机构。 该项目的目标是实现一个人工智能（AI）系统，用于识别和量化睡眠期间的呼吸事件，基于常规多导睡眠图期间收集的非侵入性心肺信号。 该项目的具体目标是从鼻插管气流信号中提取表征个体呼吸的阻力/可吸入性状态的特征。 这些将被用作神经网络的输入，该神经网络将针对已经通过上气道阻力的参考测量分类的呼吸进行训练和评估。 由于在“呼吸事件”期间通常存在“异常”呼吸的集群，因此研究者提出结合来自个体呼吸的这种分类的信息，以使用经训练的神经网络仅基于流量信号来检测和分类呼吸事件。 还将评价心肺信号在这些事件的检测和分类中的实用性。

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The objective of this project is to implement an artificial intelligence (AI) system for the identification and quantification of the respiratory events during sleep, based on noninvasive cardiopulmonary signals collected during routine polysomnography. The specific aims of the project are to extract features from the nasal cannula airflow signal that characterize the state of resistance/collapsibility of individual breaths. These will be used as inputs to a neural network that will be trained and evaluated against breaths that have been classified by reference measurement of upper airway resistance. Since clusters of "abnormal" breaths are generally present during "respiratory events," the investigator proposes to incorporate information from this classification of individual breaths to detect and classify respiratory events based on flow signal alone using a trained neural network. Utility of cardiopulmonary signals in the detection and classification of these events will also be evaluated.