Since 1993 we have performed over 2500 clinical patient studies under our Rb-82 rest-stress cardiac protocol, and we have archived all patient demographics, scan data, and processing information. We have taken a second look at some patients' transmission scans in an effort to improve our understanding of our patients' attenuation patterns. A database of transmission scans could help to answer questions about our patient population's distribution of thorax shapes, sizes, and attenuation coefficients (AC). This research was motivated by a desire to answer questions relating to the problem of reducing the time we need to perform the transmission scan.
Reduction of transmission scan time will lead to greater patient comfort, less patient motion, and greater throughput. Our protocol requires about 45 minutes to complete, of which the transmission scan makes up about 20 minutes. To reduce this time (and improve the SNR of our measured attenuation correction) we have considered the use of segmented attenuation correction (SAC), histogram smoothing methods, and other forms of adaptive and non-linear filtering. However, we have always had some nagging questions about these methods:
The answers to the above questions are beyond the scope of this paper, but to help in answering these questions (among many others) and gain a better understanding of the problems associated with measured attenuation correction we have assembled a database of 167 patient thorax transmission scans and written software to build thorax models from the patient data. This database will give us a better idea of the range of thorax shapes, sizes and AC in our patient population and also provide us with a realistic set of models to evaluate the efficacy of the different filtering methods.
_______________________________
Last updated 6/20/95