Assessing an AI enabled solution for Lung Cancer Screening using the Digital Technology Assessment Criteria (DTAC)

Unmet-Need -Lung cancer is one of the most common cancers in the UK.

Around 73% of patients are first diagnosed at an advanced stage with only 27% of patients at early stage[1].NHS-Long-Term-Plan sets out the ambition to diagnose 75% of all cancers at an early stage by 2028[2]. -Following various successful TLHC-pilots, the-UK-NSC in 2022 recommended the introduction of a UK-wide-lung-cancer-screening-programme.

This is set to generate a huge volume of scans which would put further pressure on the UK s short-staffed radiologist-workforce. -According to researchers from Wythenshawe hospital, only 5% of patients attending the TLHC are sent to the cancer clinic for suspicious lung cancer,15% require a follow-up-chest-CT-scan in 3 months and 80% feature no abnormalities and require a follow-up-scan in 2-years[3].There is therefore an urgent need to deploy AI technology that will facilitate the delivery of such complex screening programme.

Description-of-the-innovation's-evidence-accumulated-to-date InferRead CT Lung; -can automatically detect, measure, classify and track growth of pulmonary lung nodules in chest-CT-scans. -was trained on hundreds-of-thousands of scans to ensure accuracy, robustness, and generalisability and validated through our prospective, multi-reader-multi-case studies. -is proven to reduce up to 30% exam-reading-time and up to 35% missed-nodules-for-radiologists[4][5]. -could help to reduce the workload-of-radiologists and reduce unnecessary invasive procedures thereby reducing healthcare costs. -use aligns with NHS-priorities by supporting early-detection-and-diagnosis-of-lung-cancer, which can lead to improved-patient-outcomes and decreased-healthcare-costs.

InferRead CT Lung regulatory status globally; CE Class IIa,FDA,PMDA and NMPA Class III, ISO13485:2016 and ISO9001:2015 20+ peer-reviewed-publications-in-scientific-journals and Aunt Minne “Best-New-Radiology-Software-2020” award. It's available at full-scale-commercial-level;TRL9 and deployed in 25+ hospitals across Europe and 250+ worldwide.

Planned-PPI InferRead has been developed from a co-design model with PPI-collaboration-engaging-patients-clinicians and wider-stakeholders.

Following completion of this project, we plan to work with VOCAL-at-MUNHSFT on a qualitative-evaluation-of the views-of-patients, families and the general-public on using AI-based technology for lung-nodule-detection-on-chest-CT-scans.