Lung cancer screening is expanding across Europe and internationally, but CT acquisition protocols are not applied consistently. A worldwide survey through the SOLACE consortium examined institutional and technical practices in lung cancer screening and found marked variation in how centres define, update and use CT protocols. Differences were reported in protocol governance, device use, acquisition and reconstruction settings, and software support. Because lung nodule detection, volumetry and volume doubling time rely on reproducible imaging, this variation creates challenges for standardisation, quality assurance and guideline development.
Protocol Governance Remains Inconsistently Shared
The survey found that responsibility for CT protocol establishment and modification often sits with a single professional group rather than a broader multidisciplinary structure. Radiologists dominated protocol establishment, accounting for 64 of 115 responses, including both lead radiologists and radiologists. In 47 of 71 institutions, only one profession was responsible for establishing protocols, and this was most often a radiologist. A similar pattern appeared in protocol modification. Among institutions where modification was possible, 40 of 60 relied on only one professional group.
Other roles appeared much less frequently. Some institutions identified internal medical physicists, radiographers, manufacturers or external medical physicists as being involved, while a small number named screening programme directors, a pulmonologist or a biomedical engineer. Eleven institutions reported that protocols could not be modified once established, reflecting settings where fixed protocols may be required.
Protocol update schedules were equally uneven. Most institutions reported making updates as needed, while others reported no updates, yearly updates, updates less than yearly or updates more than yearly. Fourteen responses indicated no updates, including some centres where changes were still possible. Taken together, these findings indicate that protocol governance is often narrow in professional scope and irregular in timing, which complicates the development of harmonised practice.
Must Read: Automated CT Radiomics for HGSOC Prognosis
Technical Parameters Vary Across Screening Centres
The technical results revealed clear heterogeneity in equipment, reconstruction choices and image parameters. The 71 institutions reported using one CT device, two devices or more than two devices for screening, and the range of named systems covered multiple manufacturers, with many devices represented only once. High-end systems appeared more frequently, but the overall device landscape remained diverse.
Most institutions that answered the relevant question reported using automatic exposure control, indicating that dose management through automation is common in lung cancer screening. Even so, answers on detector configuration showed that some respondents misunderstood the question or interpreted it differently, suggesting uneven familiarity with technical terminology.
Reconstructed slice thickness ranged from 0.625 mm to 1.50 mm, with 1.00 mm reported most often. The interval between reconstructed images ranged from 0.50 mm to 1.25 mm and showed greater variation than slice thickness. In general, the interval was equal to or smaller than the reconstructed slice thickness, ensuring contiguous or overlapping image reconstruction.
Reconstruction kernels also varied greatly. In reconstruction algorithm type, iterative reconstruction with statistical modelling was dominant, while deep learning-supported iterative reconstruction was also in use and filtered-back projection had become relatively uncommon. The survey therefore showed that screening centres are working with a broad mix of acquisition and reconstruction settings, which creates challenges for consistency across programmes and jurisdictions.
Software Use Expands While Harmonisation Lags
Software support formed a major part of current screening practice. Among respondents who answered that section, 90% confirmed using software functions to support lung cancer screening. The most common uses were nodule detection, nodule volumetry and calculation of volume doubling time. Smaller proportions reported using software for aid in diagnosis and for structured reporting, while a few reported no software use at all. The survey clarified that these figures describe reported utilisation only and do not address software performance or clinical effectiveness.
The broader interpretation of the findings points to a screening environment that is advancing technologically but still lacks sufficient alignment. The move from filtered-back projection towards iterative reconstruction, including deep learning-supported reconstruction, reflects ongoing technical change. At the same time, the survey highlighted concerns around variability in reconstruction kernels and around inconsistent understanding of certain technical parameters. It also noted that deep learning-based reconstruction can alter image characteristics depending on the applied reconstruction kernel, raising questions about how reconstructed images interact with downstream software for detection, volumetry and volume doubling time assessment.
The findings also showed that regional regulation and legal requirements shape protocol choices. Differences between jurisdictions affect acceptable slice thicknesses, radiation dose limits and software requirements. Global interest in standardisation was evident from the geographical spread of responses, but the survey made clear that harmonisation must work across different regulatory and resource settings. These conditions make flexible but standardised frameworks increasingly important, especially as CT screening evolves and new device capabilities continue to emerge.
The survey provides an international snapshot of how CT acquisition protocols are currently managed in lung cancer screening and where the main barriers to consistency remain. Variation appears in protocol governance, update frequency, device mix, reconstruction choices, image parameters and software use. Multiprofessional involvement in defining and maintaining protocols remains limited in many centres, while technical heterogeneity continues to complicate reproducibility and guideline development. At the same time, widespread use of automatic exposure control and growing use of software functions show that screening practice is moving forward in important areas. The overall picture is one of strong international engagement paired with uneven implementation. Harmonisation, clearer technical understanding and more consistent quality assurance remain central to improving screening practice across diverse healthcare environments.
Source: Insights into Imaging
Image Credit: iStock
References:
Konrad MFG, Nischwitz E, Chorostowska-Wynimko J et al. (2026) CT acquisition protocols in lung cancer screening: implications for guideline development from a worldwide survey. Insights Imaging; 17, 70.
