Using Deep Learning and Radiomics to Predict Lymph Node Spread in Early Esophageal Squamous Cell Carcinoma

What this Trial Is About

Researchers are developing a predictive model using deep learning and radiomics to assess the risk of lymph node metastasis in patients with early-stage esophageal squamous cell carcinoma ESCC. Lymph node metastasis is crucial in deciding treatment plans and predicting outcomes for ESCC patients. This study aims to create a non-invasive tool to help doctors make more accurate treatment decisions and improve patient care by enabling earlier, more personalized interventions. The study involves analyzing medical imaging data from a total of 500 patients with early-stage ESCC. Four hundred patients from one center are divided into training and test groups to develop and validate the model, while one hundred patients from another center are used for external validation. The predictive model is assessed by measuring its ability to accurately predict lymph node metastasis using the area under the curve AUC and accuracy ACC values. Participants provide preoperative contrast-enhanced CT scans taken within two weeks before surgery, and their pathological results are reviewed. The models performance is tested on separate patient groups to ensure reliability. The primary outcome measured is the AUC value of the model over four years, which reflects how well the model can differentiate between patients with and without lymph node metastasis. The study is observational and does not involve treatment interventions or changes to usual care.

Lymph Node Metastasis in Early Esophageal Squamous Cell Carcinoma