Keywords:
Thorax, Respiratory system, CT, Computer Applications-Detection, diagnosis, Technology assessment, Drugs / Reactions
Authors:
N. R. Bogot, I. Hadas-Halpern, J. Cohen, A. Rokach, G. Izbicki, O. Benjaminov, N. Arish; Jerusalem/IL
DOI:
10.26044/ecr2019/C-3047
Methods and materials
Following IRB approval,
83 patients (mean age 76 range 58-95; male 48/58%) with suspected AILD,
treated with ≥200 mg/d of amiodarone for ≥4 months,
underwent clinically indicated DECT (Siemens SomatomFlash Scanner) in DE mode (100 and Sn140SkV) without IV contrast material.
Two radiologists with experience in DECT evaluated the CT studies independently.
Lung parenchymal disease and liver amiodarone iodine deposition were identified using dedicated DE post-processing application (Syngo.DualEnergy,
Siemens).
AILD probability was assessed on a 3-point probability scale (low,
intermediate,
high).
In addition,
two independent pulmonologists assessed AILD clinical probability on 3-point scale on the basis of patients symptoms,
dyspnea questionnaires,
pulmonary function tests (PFT) and standard CT.
Interclass correlation coefficient (ICC),
including type and absolute agreement were computed in order to assess the imaging/clinical diagnoses correlation among radiologists,
pulmonologists and between the two groups of professionals.
Read more:
In order to asses AILD diagnosis on DECT,
the software used three material decomposition algorithms of iodine and soft tissue based on characteristic changes of attenuation at 100 and 140 kV.
Iodine distribution was displayed on color coded images superimposed on regular gray scale images,
and were blended to display standard morphological information or iodine specific images.
Areas of parenchymal lung disease including air space opacities,
areas of atelectasis,
ground glass opacities and lung nodules as well as liver parenchyma were evaluated for the presence of iodine deposits on iodine specific images.