The National Lung Cancer Roundtable (NLCRT) 2020 annual meeting was the latest in a series of conferences that took place virtually in 2020.
NLCRT is an American Cancer Society (ACS) initiative which brings the nation’s top physicians and opinion leaders across all medical specialties together with leading government, non-profit, and industry organizations committed to fighting lung cancer. A packed conference agenda covered the latest developments in reducing the incidence and mortality of lung cancer.
One presentation of particular interest to our followers came from Dr Anil Vachani MD, Co-Director of Lung Cancer Screening and Associate Professor of Medicine at the Hospital of the University of Pennsylvania and the Veteran’s Administration Medical Center. Dr Vachani presented a virtual poster as a three-minute video as part of a session that highlighted lung cancer projects across the continuum of care.
The poster was titled Imaging AI/”Radiomics” decision support improves physicians’ stratification of indeterminate pulmonary nodules: A Multiple-Reader Multiple-Case study and presented a landmark study demonstrating the capabilities of AI assisted diagnosis using clinical decision support software developed by Optellum.
For this study a dataset was generated of 300 cases of solid and semi solid lung nodules, 150 of which were cancerous and 150 were benign. 12 readers, 6 pulmonologists and 6 radiologists read each case, firstly without the AI scoring and secondly assisted by the AI, to estimate a likelihood of malignancy and make a clinical recommendation. Overall, the assessment of malignancy significantly improved with the AI assistance, and the readers showed improvement in both sensitivity and specificity in determining the risk of a lung nodule being cancerous. In this study, the data showed that the lung cancer prediction digital biomarker helped improve 26% of clinical decisions made by the readers.
Dr Vachani commented afterwards:
“Clinical management of lung nodules is a challenge, particularly those incidental nodules identified independently of a screening program. Innovations such as the Optellum imaging AI will assist physicians to manage a lung nodule population and make confident decisions resulting in better outcomes.”
Download the poster here.