Development of low-dose protocols for thin-section CT assessment of cystic fibrosis in pediatric patients

Purpose: To develop low-dose thin-section computed tomographic (CT) protocols for assessment of cystic fibrosis (CF) in pediatric patients and determine the clinical usefulness thereof compared with chest radiography. Materials and Methods: After institutional review board approval and informed consent from patients or guardians were obtained, 14 patients with CF and 11 patients without CF (16 male, nine female; mean age, 12.6 years ± 5.4 [standard deviation]; range, 3.5-25 years) who underwent imaging for clinical reasons underwent low-dose thin-section CT. Sections 1 mm thick (protocol A) were used in 10 patients, and sections 0.5 mm thick (protocol B) were used in 15 patients at six levels at 120 kVp and 30-50 mA. Image quality and diagnostic acceptability were scored qualitatively and quantitatively by two radiologists who also quantified disease severity at thin-section CT and chest radiography. Effective doses were calculated by using a CT dosimetry calculator. Results: Low-dose thin-section CT was performed with mean effective doses of 0.19 mSv ± 0.03 for protocol A and 0.14 mSv ± 0.04 for protocol B (P < .005). Diagnostic acceptability and depiction of bronchovascular structures at lung window settings were graded as almost excellent for both protocols, but protocol B was inferior to protocol A for mediastinal assessment (P < .02). Patients with CF had moderate lung disease with a mean Bhalla score of 9.2 ± 5.3 (range, 0-19), compared with that of patients without CF (1.1 ± 1.4; P < .001). There was excellent correlation between thin-section CT and chest radiography (r = 0.88-0.92; P < .001). Conclusion: Low-dose thin-section CT can be performed at lower effective doses than can standard CT, approaching those of chest radiography. Low-dose thin-section CT could be appropriate for evaluating bronchiectasis in pediatric patients, yielding appropriate information about lung parenchyma and bronchovascular structures.