Improvement of accuracy in ultrasonic measurement of luminal surface roughness of carotid arterial wall by deconvolution filtering

The diagnosis of early-stage atherosclerosis is important for preventing cardiovascular diseases such as a stroke or a heart attack. The main cause of such diseases is atherosclerosis. In early-stage atherosclerosis, the luminal surface of the arterial wall becomes rough because of the detachment of endothelial cells and the degeneration of the internal elastic layer. Therefore, it would be useful if the minute luminal surface roughness of the carotid arterial wall observed in the early stage of atherosclerosis can be measured noninvasively by ultrasonography. In conventional ultrasonography, the axial resolution of a B-mode image depends on the ultrasonic pulse length of 150 mu m (at 10 MHz) because a B-mode image is constructed using the amplitude of the RF echo. However, the thickness of an endothelial cell covering the luminal surface is 10-20 mu m. Therefore, a minute surface roughness cannot be measured from a conventional B-mode image. We have realized the transcutaneous measurement of such a minute surface roughness of about 10 mu m using the phased-tracking method. However, the lateral spatial resolution degrades owing to the point spread function (PSF) because the diameter of an ultrasonic beam is finite. In the present study, we proposed a method of improving the lateral spatial resolution in the estimation of surface roughness using a Wiener filter to suppress the effect of the PSF of the ultrasound system employed. The proposed method was validated using a phantom and then applied to the in vivo measurement of carotid arteries. (C) 2014 The Japan Society of Applied Physics