This study proposes to establish a CT radiomics-based prediction model for identifying metastasis lymph nodes in esophageal cancer.

This study proposes to establish an image feature extraction and selection method for identifying lymph node metastasis of esophageal cancer.The study will consider the relationship between the main lesion and the suspected lymph node, extract association characteristics in order to improve the sensitivity and specificity of the CT radiomics-based prediction model for identifying metastasis lymph nodes. The study includes the construction of CT radiomics-based prediction model and the validation of the prediction model.

Inclusion Criteria:

• pathologically proven esophageal cancer
• scheduled to preoperative enhanced chest CT and undergo surgery
• sign the patient informed consent

Exclusion Criteria:

• failed to receive preoperative enhanced chest CT or undergo surgery
• inavailable pathological results for local lymph node status
• inquality of CT images for feature extraction
• patient quit

pathologically proven esophageal cancer patients who receive preoperative CT and surgical resection in our hospital.