Radiomics-based features for the recognition of microscopic vessel invasion in stage I non-small cell lung cancer
Principal Investigator
Name
Lin Deng
Degrees
M.D.
Institution
Jinshan Hospital of Fudan University
Position Title
Physician
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-752
Initial CDAS Request Approval
Jan 27, 2021
Title
Radiomics-based features for the recognition of microscopic vessel invasion in stage I non-small cell lung cancer
Summary
Microscopic vessel invasion (MVI) is an independent risk factor that affects the prognosis of stage I non-small cell lung caner (NSCLC). However, currently MVI can only be detected by postoperative pathology with a low efficiency and accuracy. Therefore, accurate preoperative evaluation of MVI in stage I NSCLC is an urgent demand for clinicians. Radiomics can further explore and quantify the rich information hidden in CT images without being interfered by tumor heterogeneity. We hypothesize the application of radiomic features extracted from pulmonary nodules and perinodular parenchyma could accurately detect MVI in stage I NSCLC. This project aims to retrospectively analyze and detect MVI, and radiomics model were established to recognize MVI in stage I NSCLC.
Aims
1. Segment the tumor and intratumoral blood vessel based on the method of deep convolutional neural network.
2. Extract and select highly correlated features of MVI in stage I NSCLC and establish corresponding radiomics model.
3. Analyze and verify the diagnostic efficiency of the radiomics model and demonstrate the diagnostic value of CT radiomics in MVI.
Collaborators
1. Jinwei Qiang Jinshan Hospital of Fudan University
2. Ying Li Jinshan Hospital of Fudan University
3. Li Wang Jinshan Hospital of Fudan University