(Sunday, 7/12/2020) [Eastern Time (GMT-4)]
Room: AAPM ePoster Library
Purpose:
This study is designed to further investigate the effect of patient size on the image quality associated with ATCM in chest CT scans.
Methods:
145 chest CT sacns were performed with ATCM.Water equivalent diameter (Dw), was chosen as patient size indicator because it was the dose correction factors that related to patient size. Patients were classed into three groups,the patients whose Dw were below the 25th percentile of the Dw distribution were considered to be small, while patients whose Dw above the 75th percentile were classified as large. And patients with Dw ranging from the 25th percentile to the 75th percentile were considered to be medium . One professional radiologist was instructed to place a circular region of interest (ROI) of 4 cm2 in a homogeneous area of the heart. Quantitative image noise was defined as the standard deviation of the average CT number (in Hounsfield units) within this ROI.Subjective image quality was assessed for the entire chest CT scan. Two radiologists independently compared CT images for diagnostic acceptability, using a 5-point scale.
Results:
The image noise were 7.94±0.46, 8.83±0.59, 9.45±0.90 for small, medium and large size groups respectively,it had positive correlations with Dw (r =0.80, P < .001). For the subjective assessment of image quality, the ? coefficient was 0.615 for diagnostic acceptability (P < .001) which indicates substantial agreement. None of the patients was assigned a grade indicating unacceptable or minimally acceptable diagnostic acceptability. For different-sized patients, the median scores were greater than 4 out of 5 and the diagnostic acceptability differences were not significant (P = .141, P=.165).
Conclusion:
The methods and results delineated in this study may aid radiologists in making decisions concerning image utilization and provide guidance for the design and optimization of chest CT protocols using ATCM.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
