| Autor: |
D'Errico L; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada.; Department of Radiology, Royal Papworth Hospital NHS Foundation Trust, Papworth Road, Cambridge Biomedical Campus, Cambridge CB2 0AY, UK., Ghali A; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada.; Department of Radiology, Aintree University Hospital, Liverpool University Hospitals NHS Foundation Trust, Lower Lane, Merseyside, Liverpool L9 7AL, UK., Pakkal M; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada., McInnis M; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada., Mehrez H; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada.; Canon Medical Systems Canada Limited, 75 Tiverton Court, Markham, ON L3R 4M8, Canada., Schuh AC; Princess Margaret Cancer Centre, 610 University Ave, Toronto, ON M5G 2C1, Canada., Kuruvilla JG; Princess Margaret Cancer Centre, 610 University Ave, Toronto, ON M5G 2C1, Canada., Minden M; Princess Margaret Cancer Centre, 610 University Ave, Toronto, ON M5G 2C1, Canada., Paul NS; Department of Medical Imaging, Toronto General Hospital, 200 Elizabeth St., Toronto, ON M5G 2C4, Canada.; Department of Medical Imaging, London Health Sciences Centre and St. Joseph's Hospital, Western University, 339 Windermere Road, London, ON N6A 5A6, Canada. |
| Abstrakt: |
Purpose: To compare the accuracy of ultra-low-dose (uLDCT) to standard-of-care low-dose chest CT (LDCT) in the detection of fungal infection in immunocompromised (IC) patients. Method and Materials: One hundred IC patients had paired chest CT scans performed with LDCT followed by uLDCT. The images were independently reviewed by three chest radiologists who assessed the image quality (IQ), diagnostic confidence, and detection of major (macro nodules, halo sign, cavitation, consolidation) and minor (4-10 mm nodules, ground-glass opacity) criteria for fungal disease using a five-point Likert score. Discrepant findings were adjudicated by a fourth chest radiologist. Box-whisker plots were used to analyze IQ and diagnostic confidence. Inter-rater reliability was assessed using interclass correlation coefficients (ICCs). The statistical difference between LDCT and uLDCT results was assessed using Wilcoxon paired test. Results: Lung reconstructions had IQ and diagnostic confidence scores (mean ± std) of 4.52 ± 0.47 and 4.63 ± 0.51 for LDCT and 3.85 ± 0.77 and 4.01 ± 0.88 for uLDCT. The images were clinically acceptable except for uLDCT in obese patients (BMI ≥ 30 kg/m 2 ), which had an IQ ranking from poor to excellent (scores 1 to 5). The accuracy in detecting major and minor radiological findings with uLDCT was 96% and 84% for all the patients. The inter-rater agreements were either moderate, good, or excellent, with ICC values of 0.51-0.96. There was no significant statistical difference between the uLDCT and LDCT ICC values ( p = 0.25). The effective dose for uLDCT was one quarter that of LDCT (CTDI vol = 0.9 mGy vs. 3.7 mGy). Conclusions: Thoracic uLDCT, at a 75% dose reduction, can replace LDCT for the detection of fungal disease in IC patients with BMI < 30.0 kg/m 2 . |
