Kidney stone localization in vitro using multiple-signal-classification.

Abstract

During shock wave lithotripsy, respiration and patient movement result in motion of a kidney stone. It has been estimated that 50% of the shock waves miss the stone. The misshots result in damage to the tissue with no therapeutic benefit. Here we employed a ring array of seven piezo-electric elements with center frequency of ∼600 kHz to track stones in vitro. Each element was used to transmit in turn, and the individual waveforms received on all elements were recorded. The Fourier transform of these data gives the multistatic response matrix, whose largest eigenvalues represent the dominant scatterers in the medium. The associated eigenvectors were employed by the multiple-signal-classification (MUSIC) method to generate a scattering indicator image from which scatterer locations were determined. We tested MUSIC’s applicability to track artificial kidney stones in water or with a tissue phantom. Single targets were tracked with an accuracy of 4 mm. A single stone resulted in three eigenvalues and therefore the seven element array could track two targets. Two targets were resolved for separations of 10 mm. The results indicate that MUSIC can be employed to track kidney stones although more elements would be required to track multiple fragments. [Work supported in part by NIH DK-43881].