Zhixia Su

M.D.

Yangzhou University

Not application.

PLCO (Learn more about this study)

PLCO-1498

Mar 25, 2024

A metabolomics-based multidimensional deep learning model for predicting the risk of malignancy in pulmonary nodules

Lung cancer is the second most common cancer and the leading cause of cancer-related deaths globally. Approximately half of lung cancer patients are diagnosed at an advanced stage, missing the optimal treatment window, which contributes to a significant mortality rate. Therefore, the early detection and diagnosis of lung cancer are of great public health and clinical importance. Chest X-ray and sputum cytology have been utilized for lung cancer screening since the 1970s. However, the sensitivity of these modalities was low. Over the past decades, the US National Lung Cancer Screening Trial (NLST) and several other trials in Europe have demonstrated that low-dose computed tomography (LDCT) screening effectively reduces lung cancer mortality by enabling earlier stage diagnosis. Screening for lung cancer using LDCT has led to a significant rise in the detection of solitary pulmonary nodules (PNs) in asymptomatic individuals. However, only a small fraction of these nodules are actually malignant. Therefore, it is essential to continue developing efficient and easy-to-use techniques to distinguish between benign and malignant PNs, as this is crucial for guiding clinical decision-making.
With the rapid development of histological technologies (including genomics, transcriptomics, proteomics and metabolomics), researchers have gained a more comprehensive knowledge and understanding of early screening for lung cancer. In this study, we integrated the multi-dimensional information of metabolomics and imaging, and used machine learning algorithms to construct a prediction model of malignant risk of lung nodules, which can provide a reference for the identification of malignant lung nodules and early lung cancer.