Detectability of Lung Nodules in Ultra-low Dose CT

Anticancer Res. 2021 Oct;41(10):5053-5058. doi: 10.21873/anticanres.15320.

Abstract

Background/aim: Investigation of the influence of different ultra-low dose computed tomography (ULDCT) protocols on the detection of solid and subsolid nodules in a phantom study.

Patients and methods: A chest phantom with pulmonary nodules was scanned with different CT protocols ranging from ultra-low dose settings with spectral shaping to a standard low dose lung cancer screening protocol. Image analysis was performed with different reconstruction algorithms and dedicated computer aided detection (CAD), which was compared to manual readout.

Results: The highest sensitivity rates (83%) were achieved for the 90 mAs and 120 mAs protocols when reconstructed with ADMIRE 3 or 5 and manual readout. The only statistically significant difference was found for subsolid nodules with preference of manual readout compared to CAD (p<0.05). Dose levels for the mAs settings ranged from 0.029 to 0.2 mSv.

Conclusion: Reliable detectability rates for solid nodules were achieved; CAD software did not prove reliable for subsolid nodules.

Keywords: Computed tomography; computer-aided detection; detectability; iterative reconstruction; pulmonary nodules; solid; subsolid; ultra-low dose.

MeSH terms

  • Algorithms
  • Early Detection of Cancer / methods*
  • Follow-Up Studies
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Lung Neoplasms / diagnosis*
  • Lung Neoplasms / diagnostic imaging
  • Multiple Pulmonary Nodules / diagnosis*
  • Multiple Pulmonary Nodules / diagnostic imaging
  • Phantoms, Imaging*
  • Prognosis
  • Prospective Studies
  • Radiation Dosage
  • Radiographic Image Interpretation, Computer-Assisted / methods*
  • Tomography, X-Ray Computed / methods*