ABSTRACT
This chapter deals with the use of deconvolution to improve the calculation of shunts from first pass radionuclide angiocardiography. The Wiener deconvolution filter uses a priori statistical information about the signal and the noise to improve the deconvolution. In addition to the Wiener filter, the chapter describes other methods of including a priori information into the deconvolution process. These methods can greatly improve the accuracy of the deconvolution. The first use of radionuclides in medicine was for radionuclide angiocardiography. The next use of radionuclides was by Prinzmetal in 1949. He described the radiocardiogram — the time activity curve measured from the region of the heart. Calculation of the left to right shunt ratio from a radionuclide angiocardiogram depends upon a good bolus injection. The chapter describes a linear, nontime-invariant model for first pass radionuclide angiocardiography. Wiener deconvolution requires knowledge about the power spectral densities of the system, Sa(ω), and the noise, Sn(ω).
