ntroduction- Left ventricular ejection fraction (LVEF) is an important clinical index in diagnosis and treatment of myocardial diseases. There are three major methods for measurement of LVEF: echocardiography, angiography and ECG-gated SPECT. The first method is economical, safe and rapid. The second one is more accurate, however invasive and the third one lies somewhere in between. Multi-gated SPECT usually suffers from low reproducibility compared to other methods. This is due to variation in processing factors, mainly reconstruction methods and filtration. Up to now, there is no standard method of processing of cardiac SPECT.
Methods- In this study, we attempted to find the optimum processing protocol in which the results are consistent with the angiographic results. Forty patients (referred to our department for myocardial perfusion SPECT) who had angiography within 2 weeks of the test were included in the study. All the patients had a positive history of myocardial infarction. All imaging performed with a single head GE gamma camera SPECT system model DSX using ^99mTc- MIBI. Two commonly used reconstruction methods i.e., filtered back projection and ordered subset expectation maximization with different parameters were used and the results compared with that of echocardiography and angiography.
Results- In filtered back projection technique, maximum correlation between ECG-Gated SPECT and angiography (r = 0.775) was observed when using Metz filter with psf FWHM=5 and order=5. In OSEM reconstruction technique, the maximum correlation (r = 0.706) between ECG-Gated SPECT and angiography was found using iteration of 2 and subset of 12.
Discussion- Angiography is usually assumed the standard method for calculation of LVEF.
However, techniques such as ECG-gated SPECT can be equally accurate if the method of processing is selected optimally. We optimized the method of processing in our department in correlation with the results of angiography.