Application of iDose4 Iterative Reconstruction on Low Dose CT Scanning in AIDS Patients with PJP

Objective:To study the efficacy of different iterative scales of iDose4 reconstruction algorithm in low dose chest CT scanning in AIDS patients with PJP. Methods:Sixteen AIDS patients with PJP underwent routine dose and low dose unenhanced chest CT scanning on Philips iCT. The DRI index of routine dose group and low dose group was 30 and 10, respectively. The tube voltage was 120 kV in two group. The reconstruction algorithms of routine dose group and low dose group were FPB and iDose 4(L2, L4, L6), respectively. The objective noise value, the CT dose index volume(CTDI) and dose length product(DLP) were recorded, while the effective dose(ED) was estimated. The radiation dose, the objective index of the image(image noise) and the subjective index(image quality subjective score) were compared. Results:The mean values of objective noise of LD_iDose4_L2 group, LD_iDose4_L4 group and LD_iDose4_L6 image were 12.41, 10.13 and 8.01, respectively. There was no significant difference in image noise scores between LD_iDose4_L6 group and SD_FPB group. For the lung images, the noise score of the CT image in the LD-iDose4_L2 group and the LD-iDose4_L4 group was more than 3 points, which met the needs of diagnosis. The LD_iDose4_L2 group showed the best image quality. For the mediastinum images, the subjective scores of all groups in the low-dose group were more than 3 points, which met the needs of diagnosis. The effective dose of low-dose group and the routine dose group were 6.18 and 2.28 respectively(P<0.05). The effective dose of the low dose group was reduced by 63.11% compared with the routine dose group. Conclusion:IDose 4 iteration reconstruction has obvious advantages in displaying chest CT lesions and anatomical structures in AIDS patients with PJP, which can provide CT images to meet diagnostic needs while dramatically reducing radiation dose. Under low-dose scanning conditions, the higher the level of iDose 4 iterative reconstruction iteration, the stronger the ability of noise reduction and the smaller the image noise value. But as far as the overall image quality is concerned, LD_iDose 4_L2 group shows the best image quality in lung window view and LD_iDose 4_L6 group shows the best image quality in mediastinal window view.