Background: Tomographic myocardial imaging is widely used in the diagnosis and evaluation of patients with coronary artery disease. However, its specificity remains suboptimal because of attenuation, resolution, and motion artifacts. The purpose of this study was to optimize and assess the value of attenuation, blur, and motion correction of myocardial single photon emission computed tomographic data.
Methods and results: Forty-seven studies were selected for analysis to provide 3 patient groups. Group A consisted of 18 patients with a low likelihood of coronary artery disease who were used to construct a quantitative normal database and assess changes in the normal bull's-eye produced by filtering and by attenuation correction. Group B consisted of 13 patients with a high probability of normal results, and group C consisted of 16 patients with coronary artery disease defined on angiography. The effects of attenuation correction, especially in conjunction with RESTORE (a depth-dependent deblurring filter), have been quantitated. Analysis indicates a trend to improved sensitivity and specificity for detecting individual vessel disease in this retrospective study. The motion correction program was successfully applied to 93% of patients but detected significant motion requiring correction in only 11 (24%) patients.
Conclusion: This preliminary retrospective study indicates a potential for improved myocardial single photon emission computed tomography imaging with the use of attenuation and motion correction together with a restorative deblurring filter. Confirmation by a multicenter study and larger patient numbers remain necessary to assess fully the prospective value of the technique.