Abstract:Objective: To evaluate image quality(IQ) and effective radiation dose on 256-slice coronary CTA at 50% reduced radiation dose using iterative reconstruction(IR)(iDose4) and to assess the accuracy of this protocol in the diagnosis of coronary artery disease(CAD) in patients with atypical chest pain based on selective invasive coronary angiography as gold standard. Methods: Retrospective analysis of 132 consecutive patients(83 males, mean age 54 years, mean BMI 27) with atypical chest pain underwent low-dose 256-slice CT angiography with prospective gating(Brilliance iCT, Philips Healthcare). The tube output was reduced by 50% compared to our routine institutional protocol. Scanning conditions were as follows: 120 kVp, 105 mAs, using prospective ECG-triggering at 75% of R-R interval. The HR was stabilized and controlled to ≤70 beats/min. Images were reconstructed using IR, using iterative levels 4(IR), and standard convolution kernel(XCB). Two radiologists evaluated IQ(contrast, sharpness, subjective noise and image acceptability) using a 4-point scale in a blinded manner. Subjective score consistency, the mean of subjective score and the acceptability display of coronary segments were calculate. Of the 132 patients, 55 also underwent coronary angiography, with percent diameter stenosis ≥50% as positive, to calculate the sensitivity, specificity, accuracy, positive and negative predicative values of coronary CTA with IR. Results: Subjective score consistency was achieved(Kappa value, 0.66~0.84). The scores of contrast, sharpness, subjective noise and image acceptability were 3.97±0.17, 3.86±0.35, 3.89±0.32, 3.83±0.38, respectively. The percent of diagnostic segments in 132 patients was 98.8%(1 865/1 887). The CTDI, DLP, and ED were (8.19±0.18) mGy, (100.06±7.67) mGy·cm, and (1.40±0.11) mSv, respectively. In 55 patients underwent coronary angiography, low-dose scanning with IR exhibited high accuracy. The sensitivity, specificity, and accuracy were 93.75%, 97.03% and 97.04%, respectively(segment-based); and 96%, 93.3%, and 94.5%(patient-based). Conclusion: The iDose4 iterative reconstruction technique can maintain image quality in coronary CTA scanning at 50% radiation dose reduction, and can diagnose CAD with high accuracy.