Ranish Deedar Ali Khawaja MD, Presenter: Nothing to Disclose

Sarabjeet Singh MD, Abstract Co-Author: Research Grant, Siemens AG Research Grant, Toshiba Corporation Research Grant, General Electric Company Research Grant, Koninklijke Philips NV

Rachna Madan MD, Abstract Co-Author: Nothing to Disclose

Subba Rao Digumarthy MD, Abstract Co-Author: Nothing to Disclose

Jo-Anne O. Shepard MD, Abstract Co-Author: Consultant, Agfa-Gevaert Group

Mannudeep K. S. Kalra MD, Abstract Co-Author: Nothing to Disclose

Roberto Lo Gullo MD, Abstract Co-Author: Nothing to Disclose

Atul Padole MD, Abstract Co-Author: Nothing to Disclose

Sarvenaz Pourjabbar MD, Abstract Co-Author: Nothing to Disclose

Alexi Otrakji MD, Abstract Co-Author: Nothing to Disclose

To assess diagnostic performance of chest CT examinations reconstructed with five different iterative reconstruction techniques (IRT) from major CT vendors at lowest possible radiation doses with a comparison to standard of care CT. 

This multi-phase prospective randomized trial was HIPAA-compliant & IRB-approved. In 3 different phases, 3 separate patient cohorts underwent a routine chest CT on 3 different scanners from 3 different vendors (64 MDCT and above). In addition to standard-of-care CT (SD; CTDIvol 6 mGy), all patients (n=72; 26-87 years; M:F 39:33) provided written informed consent for acquisition of additional 3 ultra low-dose “LD” series (0.9, 0.4 & 0.2mGy) immediately after SD-exam. SD-data were reconstructed with FBP (reference-standard) and LD data were reconstructed with 5 IR (IRA, IRB, IRC, IRD & IRE) at 2.5mm thickness resulting into 417 total image series. Independent blinded evaluation of lesion detection & diagnostic quality was performed. Objective noise (HU ±SD) was measured. Student’s t-test, ANOVA and Jackknife free-response receiver operating characteristic method were used for data analysis.

242 lesions (lung, 178; mediastinal, 64) were detected on standard of care chest CT. Lesion detection for both lung & mediastinal lesions was optimal across all studied 0.9 mGy IRT images (94-100% detection rate). At 0.5 mGy, detection rate for mediastinal lesions was substantially lower for IRC, IRD and IRE (33-41% missed lesions compared to SD CT). Detection for lung lesions was substantially lower for IRA, IRB and IRC (24-36% missed lesions). At 0.2 mGy, IRE missed most abnormalities (pulmonary 58% & mediastinal 60%). IRA missed the least mediastinal lesions (20%). IRC missed the least lung lesions (15%). Subjective image quality of lung parenchyma and soft-tissues (table 1) was statistically different across IR techniques (P <0.001). 

Iterative reconstruction techniques from major CT vendors differ in lesion detection at radiation dose of 0.5 mGy & lower. Subjective image quality is also statistically different across IRT at ultra-low dose chest CT.  

Low dose CT protocols from one vendor iterative reconstruction technique cannot be applied to other vendors due to significant differences in image quality and lesion detection. 

Khawaja, R, Singh, S, Madan, R, Digumarthy, S, Shepard, J, Kalra, M, Lo Gullo, R, Padole, A, Pourjabbar, S, Otrakji, A, Five Different Iterative Reconstruction Techniques across Three Vendors at the Lowest Possible Radiation Doses in Chest CT: Prospective Randomized Blinded Study.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14004245.html