Towards optimised CT lung cancer screening scan protocols.

Iball, GR.; Porter, CA.

Towards optimised CT lung cancer screening scan protocols.

All Authors

Iball, GR.

Porter, CA.

LTHT Author

Porter, Charlotte

LTHT Department

Oncology
Medical Physics & Engineering

Non Medic

Clinical Scientist

Publication Date

2026

Item Type

Journal Article

Subject

LUNG NEOPLASMS , MASS SCREENING , DIAGNOSTIC IMAGING , TOMOGRAPHY , X-RAY COMPUTED

Abstract

OBJECTIVES: To evaluate radiation dose and image quality for CT lung cancer screening protocols and to determine which reconstruction kernels deliver accurate nodule volumetry. METHODS: A variable size anthropomorphic chest phantom and a lung screening image quality phantom were scanned on 17 different scanner models. Dose metrics from the chest phantom scans were compared against international dose recommendations.Image quality phantom scans were used to assess whether the protocol yielded accurate nodule volumetry as per the Quantitative Imaging Biomarkers Alliance (QIBA) criteria, and to determine whether changes were needed to enable this. RESULTS: Doses varied by up to a factor of seven for the largest phantom size. All protocols delivered doses below the UK standards but only two scanners met the European dose criteria. Many scanners failed to meet the QIBA criteria with lung reconstruction kernels, primarily due to excessive edge enhancement and insufficient 3-dimensional spatial resolution. With alternative reconstruction kernels, the QIBA criteria were met on all bar two scanners. CONCLUSIONS: Standard lung screening CT protocols on contemporary scanners produce images of varying image quality, over a wide range of radiation doses. Scanner-specific reconstruction kernel selection is required to meet the QIBA criteria. The study findings enable the development of optimised CT scanning protocols for lung cancer screening programmes. ADVANCES IN KNOWLEDGE: Standard lung cancer screening scanning protocols demonstrate notable variations in dose. Lung reconstruction kernels often yield inaccurate measurements of nodule volume; instead, soft tissue kernels are necessary to obtain reliable results.