Feasibility of lung cancer prediction from low-dose CT scan and smoking factors using causal models

• Published in: Thorax Vol. 74; no. 7; pp. 643 - 649
• Main Authors: Raghu, Vineet K, Zhao, Wei, Pu, Jiantao, Leader, Joseph K, Wang, Renwei, Herman, James, Yuan, Jian-Min, Benos, Panayiotis V, Wilson, David O
• Format: Journal Article
• Language: English
• Published: England BMJ Publishing Group Ltd and British Thoracic Society 01.07.2019; BMJ Publishing Group LTD; BMJ Publishing Group
• Series: Original article
• Subjects: Aged; Angiogenesis; Artificial intelligence; Bioinformatics; cancer screening; Diagnosis, Differential; Early Detection of Cancer - methods; Feasibility Studies; Female; Humans; low-dose CT; Lung Cancer; lung cancer risk; Lung Neoplasms - diagnostic imaging; Lung Neoplasms - etiology; Lung Neoplasms - pathology; Male; Mass Screening - methods; Medical imaging; Medical screening; Middle Aged; Models, Statistical; Multiple Pulmonary Nodules - diagnostic imaging; Multiple Pulmonary Nodules - pathology; Predictive Value of Tests; Radiation Dosage; Risk Factors; Smoking - adverse effects; Smoking Cessation - statistics & numerical data; Tomography; Tomography, X-Ray Computed - methods; Variables; United States > US; Pittsburgh Pennsylvania; lung cancer risk; low-dose CT; cancer screening
• ISSN: 0040-6376; 1468-3296; 1468-3296
• DOI: 10.1136/thoraxjnl-2018-212638

IntroductionLow-dose CT (LDCT) is currently used in lung cancer screening of high-risk populations for early lung cancer diagnosis. However, 96% of individuals with detected nodules are false positives.MethodsIn order to develop an efficient early lung cancer predictor from clinical, demographic and LDCT features, we studied a total of 218 subjects with lung cancer or benign nodules. Probabilistic graphical models (PGMs) were used to integrate demographics, clinical data and LDCT features from 92 subjects (training cohort) from the Pittsburgh Lung Screening Study cohort.ResultsLearnt PGMs identified three variables directly (causally) linked to malignant nodules and the largest benign nodule and used them to build the Lung Cancer Causal Model (LCCM), which was validated in a separate cohort of 126 subjects. Nodule and vessel numbers and years since the subject quit smoking were sufficient to discriminate malignant from benign nodules. Comparison with existing predictors in the training and validation cohorts showed that (1) incorporating LDCT scan features greatly enhances predictive accuracy; and (2) LCCM improves cancer detection over existing methods, including the Brock parsimonious model (p<0.001). Notably, the number of surrounding vessels, a feature not previously used in predictive models, significantly improves predictive efficiency. Based on the validation cohort results, LCCM is able to identify 30% of the benign nodules without risk of misclassifying cancer nodules.DiscussionLCCM shows promise as a lung cancer predictor as it is significantly improved over existing models. Validated in a larger, prospective study, it may help reduce unnecessary follow-up visits and procedures.