Diagnostic Performance of Motion Compensation Reconstruction Algorithm Based on Artificial Intelligence in Coronary Artery CT Angiography in Patients with Different Heart Rates
Objective To investigate the effect of motion compensation reconstruction(MCR)based on artificial intel-ligence(AI)on coronary image quality and coronary artery evaluability in different heart rate groups.Methods A total of 181 patients with suspected or diagnosed coronary heart disease who underwent coronary artery CT angiography(CCTA)with second-generation double-layer detector CT(DLCT)were prospectively included.According to the heart rate,the pa-tients were divided into three groups from low to high(heart rate≤65 beats/min,66-79 beats/min,≥80 beats/min).Each group reconstructed images with or without MCR algorithm based on artificial intelligence.The double-blind e-valuation was performed by two doctors with more than 5 years working experience in cardiovascular direction in radiology department.Subjective image quality was evaluated by 5-point method,and more than 3 points were considered as good im-age quality,while objective image quality was evaluated by signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR).Some patients underwent digital subtraction angiography(DSA)as the"gold standard"to compare the diagnostic perform-ance of conventional and superimposed MCR in the diagnosis of lumen stenosis(≥50%).Results There were 53 pa-tients in low heart rate group,79 patients in middle rate group and 49 patients in high heart rate group.There were signifi-cant differences in subjective image quality among the three groups(P<0.05),but there was no significant difference in SNR and CNR among the three groups.After adding MCR,the subjective score of the image was higher,of which 54(7.46%)were considered to be undiagnosable vessels on conventional images,while only 5(0.69%)were on MCR.MCR re-duced image noise and improved image quality,especially in center rate and high heart rate groups(P<0.001).In vascu-lar level analysis,compared with conventional MCR,superimposed MCR increased sensitivity(73.4%vs.90.6%)and specificity(94.2%vs.97.8%).Conclusion The second generation double-layer spectral detector CT can obtain satis-factory images of the heart rate of patients below 80 beats per minute.MCR technology based on artificial intelligence can improve the image quality of coronary artery,especially in patients with high heart rate,and has high consistency with DSA,which provides important information for clinical diagnosis and treatment.
NETL
NSTL
万方数据
