安秀静 1吴哲 2姜超 3李宁 3万鞠冰 2王森 3李东遥 3田卢峰1
作者信息
- 1. 锦州医科大学抚顺市中心医院研究生培养基地 抚顺 113000||抚顺市中心医院内科 抚顺 113000
- 2. 锦州医科大学抚顺市中心医院研究生培养基地 抚顺 113000||抚顺市中心医院放射科 抚顺 113000
- 3. 抚顺市中心医院放射科 抚顺 113000
- 折叠
摘要
Abstract
Objective:To explore the feasibility of using low-dose computed tomography(LDCT)with deep learning reconstruction(DLR)on the chest for the screening of lung nodules,and to compare the image quality and detection rate of nodules between LDCT and routine dose CT(RDCT)-DLR.Methods:A total of 104 overweight or obese patients[body mass index(BMI)≥25 kg/m2]who received CT examination on chest due to pulmonary nodule screening from September to December 2023 were included to conduct prospective study.All patients underwent respectively RDCT(120 kV)and LDCT(100 kV)scans,all of the two scans used the modulation of automatic tube current,and adopted deep learning AI algorithm ClearInfinity to conduct reconstruction(RDCT:CI 40%,LDCT:CI 50%).Radiation dose and nodules number of them were recorded.At the T8 vertebral level,CT values(Hounsfield Units,HU)of mediastinal fat and lung parenchyma in the right lower lobe were measured,along with image noise(standard deviation,SD).The signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were subsequently calculated.Two radiologists independently performed subjective evaluations of image quality and pulmonary nodules using a Likert 4-point scale.Paired t-tests or Wilcoxon rank-sum test were employed to compare differences in radiation dose,objective image noise,and subjective scores between LDCT and RDCT.Results:A total of 104 patients were enrolled,including 54 males and 50 females,with a mean age of 52±13 years and a BMI of(27.77±2.64)kg/m2.The effective radiation dose of LDCT demonstrated a statistically significant reduction compared to RDCT(Z=-8.853,P<0.001),with a mean effective radiation dose reduction of 77.86%.The differences in lung CT value,lung noise,fat noise,lung parenchyma SNR,fat SNR and CNR of images between two groups were significant(Z=-3.022,-2.327,-4.785,-2.059,-3.765,-4.013,P<0.05),while there were not significant differences in the comparisons for fat CT value and lung parenchyma SNR(P>0.05).The image contrast,image noise,and subjective score for image quality of lung nodule of LDCT were lower than those of RDCT(t=2.877,2.387,5.096,P<0.05),but all subjective scores of that were>3,which can meet the requirements of clinical diagnosis.In terms of nodule detection,RDCT found out about 418 nodules,while LDCT found out about 421,the false positive rate of LDCT only was 0.72%.Conclusion:In overweight or obese patients,LDCT that combined with DLR algorithm on chest is equivalent to RDCT on image quality and the detection rate of lung nodule,and it significantly reduce radiation exposure on patients at the same time.关键词
肺结节/低剂量计算机断层扫描/深度学习重建算法/图像质量Key words
Lung nodule/Low-dose computed tomography(LDCT)/Deep learning reconstruction algorithm/Image quality分类
医药卫生
