This study compares the sensitivity, specificity and accuracy of radiologists, thoracic surgeons and a predictive model (PKUM model) to discriminate malignancy from benign nodules in patients with multiple pulmonary nodules.

It is clinically difficult to diagnose and manage patients with multiple pulmonary nodules (MPNs). We have developed a web-based mathematic model (PKUM model) by using a multi-centric database from three institutions (Peking University People's Hospital, Haidian Section of Peking University Third Hospital, People's Hospital Affiliated to Hebei Medical University) to predict the probability of a nodule to be malignant in patients with MPNs. This prospectively observational study will recruit patients with MPNs between January 2019 and March 2019, allowing radiologists, surgeons, and a predictive model (PKUM model) to discriminate malignancy from benign nodules, and compare their sensitivity, specificity, and accuracy.

Inclusion Criteria:

• Patients with newly discovered, 4-30 mm multiple pulmonary nodules shown on thoracic CT scans
• Patients with at least two nodules resected for pathological evaluation

Exclusion Criteria:

• History of malignancy within 5 years
• Presence of pneumonia or pleural effusion on thoracic CT scans
• Patients with none or only one nodule resected
• Patients with initial chemo-radiation therapy

The population include patients with multiple pulmonary nodules who will be diagnosed and treated between January 1，2019 and March 30, 2019. Patients with at least 2 nodules resected will be enrolled in this study.