Investigation of digital filtering for stacked, phased ultrasound transducers

Abstract

High bandwidth transducers are of interest in all applications of ultrasound imaging. A stacked, phased multi-layer transducer was previously described that extended the bandwidth to multiple octaves. However, the frequency response of this transducer design is characterized by multiple peaks and troughs that will result in a 'ringy' waveform. However, Digital Signal Processing (DSP) and related devices are ubiquitous. Modern ultrasound systems digitize received signals and incorporate digital filters. In this paper the utility of digital filtering for improving transducer frequency response was tested in MATLAB. A Finite Element Analysis (FEA) model was developed in PZFlex to simulate a transducer array element. Our FEA and Matlab simulation results indicate that a matched filter in transducer receive path can remove the 'bumps' in the spectrum and reduce the -20dB pulse length by approximately 14%. An inverse filter employed in the transmitting path can also improve the characteristics of the transmitted signal - reducing the pulse length by 35%. A Pseudo-inverse procedure was used to design these FIR filters with a desired length. Furthermore, a design parameter iteration simulation demonstrates that this digital filtering technique can work effectively even when there is variation in the material properties and transducer element dimensions. Major parameters considered included: electro-mechanical coupling coefficient (+/-10%), dielectric permittivity (+/-10%) and transducer element thickness (+/-5%). The results demonstrate that digital filtering can play an important role in compensating for transducer spectral response imperfections. Thus, we can obtain reduced pulse length and improved imaging resolution with high bandwidth multi-layer transducers. The approach also has value for compensating for spectral imperfections in conventional single layer transducers.