A phantom study to optimise the automatic tube current modulation for chest CT in COVID-19

Area of Science:

• Radiology
• Medical Physics
• Infectious Diseases

Background:

• The COVID-19 pandemic necessitates careful use of thoracic computed tomography (CT).
• Low-dose CT protocols are recommended to minimize radiation exposure.
• Detecting early ground-glass opacities (GGO) in COVID-19 is crucial for diagnosis.

Purpose of the Study:

• To determine the automatic tube current modulation noise index threshold for detecting early COVID-19 GGO.
• To establish a safe low-dose CT protocol for COVID-19 diagnosis.

Main Methods:

• Performed 25 phantom CT studies with varying automatic tube current modulation settings (SUREExposure3D).
• Retrospectively analyzed chest CT images from 22 COVID-19 patients.
• Calculated density differences between GGO and unaffected lung tissue.
• Matched phantom and patient data to find the minimum noise index threshold.

Main Results:

• The minimum density difference for detecting COVID-19 GGO was 252 HU.
• This corresponded to a SUREExposure3D noise index of 36.
• Established a noise index threshold of 36 for Canon scanners (without iterative reconstruction).

Conclusions:

• A noise index threshold of 36 allows for significant dose reduction (up to 80% decrease in dose-length product) in low-dose CT for COVID-19.
• This threshold helps avoid missing early GGO in COVID-19 patients.
• Prospective validation of this protocol is recommended.