Fine details better visualized in denoised ultra-low-dose CT images in adults who were immunocompromised
By Elana Gotkine HealthDay Reporter
FRIDAY, March 21, 2025 (HealthDay News) — For individuals who are immunocompromised, denoised ultra-low-dose computed tomography (ULDCT) shows better accuracy than ULDCT for identifying lungs with or without pneumonia, according to a study published online March 13 in Radiology: Cardiothoracic Imaging.
Maximilliano Klug, M.D., from The Chaim Sheba Medical Center in Ramat Gan, Israel, and colleagues conducted a prospective study involving 54 adults who were immunocompromised, referred for a chest CT to evaluate for pneumonia between September 2020 and December 2022. The participants underwent two scans: normal-dose CT (120 kVp and automatic current modulation) and ULDCT (100 kVp and constant current of 10 mA). A postprocessing procedure was performed on the ULDCT images using an artificial intelligence algorithm to reduce noise. Images obtained from normal-dose CT, ULDCT, and denoised ULDCT were examined for the presence of pneumonia and associated findings.
The researchers found that 10 of the 54 participants were correctly identified as having no pneumonia; accuracy was similar for denoised ULDCT and ULDCT (100 and 96 to 98 percent, respectively). Both methods detected pneumonia and features associated with invasive fungal pneumonia, but slightly better accuracy was seen with denoised ULDCT (100 versus 91 to 98 percent). In denoised ULDCT images, fine details were better visualized: tree-in-bud pattern (accuracy, 93 versus 78 to 80 percent), interlobular septal thickening (accuracy, 78 to 83 versus 61 to 67 percent), and intralobular septal thickening (accuracy, 85 to 87 versus 0 percent).
“Our suggested method to denoise images obtained at chest ULDCT scans significantly reduced image noise and showed better identification of individuals with normal lungs and pneumonia,” the authors write.
Copyright © 2025 HealthDay. All rights reserved.
