Correlation of clinical and physical-technical image quality in chest CT: a human cadaver study applied on iterative reconstruction

BMC Medical Imaging

Research Article

Tom Van Hoof¹  Klaus Bacher¹  An De Crop¹  Katharina D’Herde¹  Hubert Thierens¹  Koenraad Verstraete²  Mathias Van Borsel²  Tom Dewaele²  Peter Smeets²  Merel Vergauwen²  Eric Achten² 

[1] Department of Basic Medical Sciences, Ghent University, Proeftuinstraat 86, B-9000, Ghent, Belgium;Department of Radiology, Ghent University Hospital, De Pintelaan 185, B-9000, Ghent, Belgium;

关键词: Chest CT;    Image quality;    Iterative reconstruction;    Human cadaver study;    Visual grading analysis;   

DOI  :  10.1186/s12880-015-0075-y

 received in 2014-12-02, accepted in 2015-08-10,  发布年份 2015

来源: Springer

【 摘 要 】

BackgroundThe first aim of this study was to evaluate the correlation between clinical and physical-technical image quality applied to different strengths of iterative reconstruction in chest CT images using Thiel cadaver acquisitions and Catphan images. The second aim was to determine the potential dose reduction of iterative reconstruction compared to conventional filtered back projection based on different clinical and physical-technical image quality parameters.MethodsClinical image quality was assessed using three Thiel embalmed human cadavers. A Catphan phantom was used to assess physical-technical image quality parameters such as noise, contrast-detail and contrast-to-noise ratio (CNR).Both Catphan and chest Thiel CT images were acquired on a multislice CT scanner at 120 kVp and 0.9 pitch. Six different refmAs settings were applied (12, 30, 60, 90, 120 and 150refmAs) and each scan was reconstructed using filtered back projection (FBP) and iterative reconstruction (SAFIRE) algorithms (1,3 and 5 strengths) using a sharp kernel, resulting in 24 image series. Four radiologists assessed the clinical image quality, using a visual grading analysis (VGA) technique based on the European Quality Criteria for Chest CT.ResultsCorrelation coefficients between clinical and physical-technical image quality varied from 0.88 to 0.92, depending on the selected physical-technical parameter. Depending on the strength of SAFIRE, the potential dose reduction based on noise, CNR and the inverse image quality figure (IQFinv) varied from 14.0 to 67.8 %, 16.0 to 71.5 % and 22.7 to 50.6 % respectively. Potential dose reduction based on clinical image quality varied from 27 to 37.4 %, depending on the strength of SAFIRE.ConclusionOur results demonstrate that noise assessments in a uniform phantom overestimate the potential dose reduction for the SAFIRE IR algorithm. Since the IQFinv based dose reduction is quite consistent with the clinical based dose reduction, an optimised contrast-detail phantom could improve the use of contrast-detail analysis for image quality assessment in chest CT imaging. In conclusion, one should be cautious to evaluate the performance of CT equipment taking into account only physical-technical parameters as noise and CNR, as this might give an incomplete representation of the actual clinical image quality performance.

【 授权许可】

CC BY   
© De Crop et al. 2015

【 图 表 】

12951_2015_155_Article_IEq76.gif

Fig. 1