Currently available screening programmes for lung cancer are limited by many challenges including low diagnostic accuracy, radiation exposure and high costs. New technologies in PET/CT scanners can allow cheaper and more sensitive exams with low radiation exposure. AI can be used to denoise LDCT to enhance the accuracy of imaging tests and build riskassessment models. This project aims to develop a new approach exploiting both these revolutionary advancements to bridge the existing gap in lung cancer screening. Patients in a high-risk population will be enrolled into two different cohorts undergoing LDCT scan and simultaneous [18F]FDG PET/CT on new-generation long axial field of view scanner (UO1) or screening with low LDCT only (UO2). AI will assist in image enhancement and interpretation and will develop a personalised risk-model guiding the following steps of clinical management, significantly improving early diagnosis of lung cancer, reducing mortality and healthcare costs.

Assessment of the potential added value of low dose [18F]FDG PET/CT in the early detection of lung cancer in the screening work-up of the high-risk population. The target population (cohort 1) for the multimodal screening programme will be identified at UO1 on the basis of the PLCOM2012 prediction risk model. All patients enrolled in cohort 1 will undergo a LDCT and a simultaneous low dose [18F]FDG PET scan. Whole body [18F]FDG PET/CT will be performed according to EANM Guidelines for tumour imaging. Briefly, patients will be instructed to fast for 6 hours and to avoid strenuous physical activity for 48 hours prior to the PET/CT scan. Diabetic patients will be instructed to fast for 4 hours prior to the scan and will be instructed on the use of medications according to the institution's protocol. Blood glucose levels will be checked, [18F]FDG will be injected intravenously, and PET/CT images will be acquired after 60 minutes (± 10 minutes). A new generation long axial field of view PET/CT scanner (Omni Legend, General Electric Healthcare, Waukesha, WI, USA) will be used for whole body imaging.

The control population (cohort 2) will be identified in UO2 on the basis of the PLCOM2012 prediction risk model as previously described for cohort 1. The LDCT scan will be performed in single deep inspiratory breath hold. No intravenous contrast will be administered. LDCT images will be assessed and interpreted by at least two experienced radiologists. LDCT will be defined as positive if at least one non-calcified lung nodule >5 mm in any diameter is detected. LDCT will be defined as negative if no clinically significant morphological alterations are detected. Non-calcified nodules of 3-5mm detected by LDCT will be reported. Lung-RADS criteria will be used to classify detected lung nodules. Any other suspicious morphological alteration will be reported. The number, size, characteristics, and location of any lesion detected will be recorded for each LDCT scan. Any abnormality suggestive of clinically significant conditions other than cancer will be also reported. Sensitivity, specificity, and accuracy will be calculated by per-patient and per-lesion analysis.

Inclusion Criteria:

• Age > 50 years
• PLCOm2012 risk prediction > 4%
• Be willing to adhere to the study intervention through [18F]FDG PET/CT or LDCT imaging
• Signed written informed consent form

Exclusion Criteria:

• Blood glucose levels >200 mg/dl,
• Ongoing pregnancy and breastfeeding
• Unwillingness to participate,
• Previous diagnosis of lung cancer,
• Previous CT scan within the last 24 months
• Concomitant severe clinical conditions and any condition that preclude the feasibility of the study