Microvascular Diagnostics, Northern Medical Physics and Clinical Engineering, Freeman Hospital, Newcastle upon Tyne, NE7 7DN, United Kingdom.
Physiol Meas. 2019 Sep 3;40(8):08NT01. doi: 10.1088/1361-6579/ab3545.
To assess the feasibility of novel photoplethysmography (PPG) Pulse Sounder/Pulse Visualizer communication techniques for alerting the presence (or absence) of peripheral arterial disease (PAD).
Proof of concept evaluation using our previously published multi-site PPG pulse data set (110 participants included; age > 40 years; 44% PAD by ankle brachial pressure index (ABPI)). Two main pulse encoding rules using the risetime as an example feature to mark each heartbeat in a 6 s analysis study window: if risetime at both great toes ⩽time threshold ('no PAD' state) then heartbeat marked with a single 5 kHz audio tone; if risetime from either great toe > threshold ('PAD') then heartbeat marked with a distinct train of 5 kHz audio tones. A simple spectrogram visual plot was also produced for each participant's audio signals. ROC curve analysis first determined an appropriate risetime threshold using the simple-to-evaluate PAD Visualizer and then re-tested at this level with PAD Sounder and classification performance summarized.
The pulse encoding rules used were easy to learn. Visualizer best communication performance at risetime 0.27 s gave specificity 80.6%, sensitivity 95.8%, accuracy 87.3%, negative predictive value 96.2%, and Kappa statistic of 0.75. Classification performance for blinded assessment with audio PAD Sounder was identical.
Substantial agreements demonstrated between Sounder/Visualizer disease techniques and the ABPI PAD reference. These simple-to-use PAD communications methods when embedded into an appropriate device platform could offer significant benefits in PAD diagnosis for a range of clinical settings, including primary care where low-cost, portable and easy-to-use diagnostics can be desirable.
评估新型光电容积脉搏波(PPG)脉搏发声器/脉搏可视化器通信技术在提示周围动脉疾病(PAD)存在(或不存在)方面的可行性。
使用我们之前发表的多站点 PPG 脉搏数据集进行概念验证评估(包括 110 名参与者;年龄>40 岁;踝肱指数(ABI)44%的 PAD)。使用上升时间作为示例特征的两种主要脉搏编码规则来标记 6s 分析研究窗口中的每个心跳:如果两个大脚趾的上升时间⩽时间阈值(“无 PAD”状态),则用单个 5kHz 音频音标记心跳;如果任一大脚趾的上升时间>阈值(“PAD”),则用独特的 5kHz 音频音标记心跳。还为每个参与者的音频信号生成了简单的声谱图可视化图。ROC 曲线分析首先使用易于评估的 PAD 可视化器确定适当的上升时间阈值,然后在该水平上使用 PAD 发声器重新测试,并总结分类性能。
使用的脉搏编码规则易于学习。可视化器在上升时间 0.27s 时的最佳通信性能具有 80.6%的特异性、95.8%的敏感性、87.3%的准确性、96.2%的阴性预测值和 0.75 的 Kappa 统计量。使用音频 PAD 发声器进行盲评估的分类性能相同。
发声器/可视化器疾病技术与 ABI PAD 参考之间显示出实质性的一致性。这些简单易用的 PAD 通信方法,如果嵌入到适当的设备平台中,可以为各种临床环境中的 PAD 诊断带来显著的益处,包括初级保健,在这些环境中,低成本、便携式和易于使用的诊断方法可能是理想的。