Chik William W B, Kosobrodov Roman, Bhaskaran Abhishek, Barry Michael Anthony Tony, Nguyen Doan Trang, Pouliopoulos Jim, Byth Karen, Sivagangabalan Gopal, Thomas Stuart P, Ross David L, McEwan Alistair, Kovoor Pramesh, Thiagalingam Aravinda
Cardiology Department, Westmead Hospital, Sydney, Australia.
Sydney Medical School, University of Sydney, Sydney, Australia.
J Cardiovasc Electrophysiol. 2015 Apr;26(4):440-447. doi: 10.1111/jce.12598. Epub 2015 Feb 14.
Steam pop is an explosive rupture of cardiac tissue caused by tissue overheating above 100 °C, resulting in steam formation, predisposing to serious complications associated with radiofrequency (RF) ablations. However, there are currently no reliable techniques to predict the occurrence of steam pops. We propose the utility of acoustic signals emitted during RF ablation as a novel method to predict steam pop formation and potentially prevent serious complications.
Radiofrequency generator parameters (power, impedance, and temperature) were temporally recorded during ablations performed in an in vitro bovine myocardial model. The acoustic system consisted of HTI-96-min hydrophone, microphone preamplifier, and sound card connected to a laptop computer. The hydrophone has the frequency range of 2 Hz to 30 kHz and nominal sensitivity in the range -240 to -165 dB. The sound was sampled at 96 kHz with 24-bit resolution. Output signal from the hydrophone was fed into the camera audio input to synchronize the video stream. An automated system was developed for the detection and analysis of acoustic events.
Nine steam pops were observed. Three distinct sounds were identified as warning signals, each indicating rapid steam formation and its release from tissue. These sounds had a broad frequency range up to 6 kHz with several spectral peaks around 2-3 kHz. Subjectively, these warning signals were perceived as separate loud clicks, a quick succession of clicks, or continuous squeaking noise. Characteristic acoustic signals were identified preceding 80% of pops occurrence. Six cardiologists were able to identify 65% of acoustic signals accurately preceding the pop. An automated system identified the characteristic warning signals in 85% of cases. The mean time from the first acoustic signal to pop occurrence was 46 ± 20 seconds. The automated system had 72.7% sensitivity and 88.9% specificity for predicting pops.
Easily identifiable characteristic acoustic emissions predictably occur before imminent steam popping during RF ablations. Such acoustic emissions can be carefully monitored during an ablation and may be useful to prevent serious complications during RF delivery.
蒸汽爆泡是心脏组织因温度超过100°C而发生的爆炸性破裂，导致蒸汽形成，易引发与射频（RF）消融相关的严重并发症。然而，目前尚无可靠技术来预测蒸汽爆泡的发生。我们提出将射频消融过程中发出的声学信号用作预测蒸汽爆泡形成并潜在预防严重并发症的新方法。
在体外牛心肌模型进行消融时，对射频发生器参数（功率、阻抗和温度）进行实时记录。声学系统由HTI - 96 - min水听器、麦克风前置放大器和连接到笔记本电脑的声卡组成。水听器的频率范围为2 Hz至30 kHz，标称灵敏度在 - 240至 - 165 dB之间。声音以96 kHz的采样率、24位分辨率进行采样。水听器的输出信号输入到相机音频输入以同步视频流。开发了一个用于检测和分析声学事件的自动化系统。
观察到9次蒸汽爆泡。识别出三种不同的声音作为警告信号，每种声音都表明蒸汽快速形成并从组织中释放。这些声音的频率范围很宽，高达6 kHz，在2 - 3 kHz左右有几个频谱峰值。主观上，这些警告信号被感知为单独的响亮咔嗒声、一连串快速的咔嗒声或连续的吱吱声。在80%的爆泡发生之前识别出了特征性声学信号。六位心脏病专家能够在爆泡发生前准确识别出65%的声学信号。一个自动化系统在85%的病例中识别出了特征性警告信号。从第一个声学信号到爆泡发生的平均时间为46±20秒。该自动化系统预测爆泡的灵敏度为72.7%，特异性为88.9%。
在射频消融过程中，即将发生蒸汽爆泡之前可预测地会出现易于识别的特征性声发射。在消融过程中仔细监测这种声发射可能有助于预防射频传递过程中的严重并发症。
