Aim: The aim of this study is to assess the validation of summed segmental scores in detection accuracy CAD from different reconstructed polar maps. Method: The summed scores (SSS, SRS and SDS) were obtained from standard filtered backprojection reconstruction (FBP) and iterative reconstruction with CT-AC that being performed on the raw emission data of 30 patients (15 male and 15 females) with known or suspected coronary artery disease who had undergone two-day stress-rest or rest-stress 99mTc-MIBI myocardial perfusion SPECT. The reconstruction algorithms and the resulting projected polar maps were used to estimate the radioactivity distribution within the left ventricular myocardium of the heart. Result: Mean ± SD for all summed scores is higher in FBP than CT-AC with SSS in FBP 32.60 ± 5.360, SRS FBP 31.60 ± 5.069 and SDS FBP 3.87 ± 3.159. Meanwhile for SSS in CT-AC is 28.27 ± 5.245, SRS CT-AC 27.20 ± 4.397 and SDS CT-AC 3.27 ± 2.449. There is statistically significant difference of SSS and SRS between FBP and CT-AC tested using Mann-Whitney test (p value = 0.002 for SSS and p-value = 0.001 for SRS). However, there is no statistically significant difference of SDS between FBP and CT-AC tested using Mann-Whitney test (p value = 0.633). Conclusion: Summed scores studies reconstructed with FBP are statistically not significantly different from AC. However, compensation for attenuation artifact is much better with iterative reconstruction CT-AC that FBP as CT-AC for SPECT MPI allows improved specificity of RCA territory compared with FBP reconstruction SPECT MPI.