弁機能不全の早期発見を目的としたウェーブレット変換による乱流解析

Abstract

Although it is important to find malfunctions of mechanical prosthetic heart valves at an earlier stage, the detection of a small thrombus or vegetation around valves is sometimes difficult by conventional clinical methods. In this study, we introduce the wavelet transform(WT)as a method for the detection of slight malfunctions of valves. The test valve(Bjork-Shiley valve, 25mm)was mounted at the aortic or mitral position of a computer-controlled hydraulic mock circulator and driven under normal or AF flow condition. The valve tested as a malfunctional model was mechanically restricted with the mobility of an occluder. Using continuous-wave ultrasound Doppler method, mean velocity signals were measured at3cm forward from the valve outlet, and were analyzed by Daubechies wavelet. As a result, comparing a thrombosed model with a normal one, there was a significant increase in power in the signal bandwidth(100Hz-1kHz)and it was remarkable during acceleration of flow. The increase in power is thought to be caused by turbulence of downstream flow. These in vitro results indicate that WT is effective for the detection of valve malfunctions.