Improvement of axial spatial resolution of ultrasound image using Wiener filter for measurement of intima-media thickness of carotid artery

Abstract

Measurement of intima-media thickness (IMT) of the carotid artery by ultrasonography is widely used for diagnosis of atherosclerosis. For accurate measurement of IMT, in the present study, a method based on the Wiener filter was developed for improvement of image spatial resolution. There are many studies on improvement of spatial resolution for elimination of the point spread function (PSF) of a system using a deconvolution filter. PSF is determined by the transfer function H(f ) (including transfer function of ultrasonic probe and tissue attenuation property). In most of developed methods, PSF is separately measured using a target, such as a fine wire in water, prior to the actual in vivo measurement. However, it is difficult to predict the subject-dependent change in PSF owing to frequencydependent attenuation by tissue. Therefore, conventional deconvolution filters do not work effectively. In the present study, the magnitude |H(f )| was estimated by averaging power spectra of RF echoes in all scan lines obtained from the actual in vivo measurement to consider frequency-dependent attenuation. Although there are many dips in the respective power spectra due to interference of echoes from many scatterers, such dips can be suppressed by averaging power spectra because locations of dips are different line by line owing to randomness of scatterer distribution. On the other hand, the phase of H(f ) was estimated beforehand from the phase of frequency spectrum Y (f ) of the transmitted ultrasound signal received by a hydrophone placed in water because attenuation by tissue does not affect the phase of the transfer function. The phase of squared Y (f ) was used as the phase of H(f ) by considering a pulse-echo measurement. From the estimated H(f ), a deconvolution filter was designed using the Wiener filter. Improvement of axial resolution was evaluated by measuring an echo from a fine wire placed in water. Axial resolution defined by the width at half maximum was improved from 0.2 mm to 0.12 mm. Furthermore, the proposed method was applied to RF echoes obtained from a carotid artery. The axial resolution was significantly improved compared with conventional B-mode image.

Denoising Ultrasound Medical Images

“…Various other techniques combine impulse removal filters with local adaptive filtering in the transform domain to remove not only white and mixed noise, but also their mixtures (Rabbani, Vafadust, Abolmaesumi, & Gazor, 2008). In order to reduce the presence of noise in medical images many techniques are available from the past such as linear filtering (Simon,VanBaren & Ebbini, 1998) adaptive filtering (Weiner Filters) (Hasegawa, Kageyama & Kanai, 2013) and median filtering (Czerwinski, Jones & O'Brien, 1995). However, digital filters, linear filters and adaptive filters proved to reduce noise in stationary signals.…”