Using Machine Learning to Predict the Prognosis of Cervical Cancer Patients with Lymph Node Metastasis: An Analysis...

Area of Science:

• Oncology
• Medical Informatics
• Machine Learning

Background:

• Cervical cancer is a leading cause of cancer death in women globally.
• Lymph node metastasis significantly worsens prognosis in cervical cancer patients.
• Accurate prognostic prediction is crucial for effective clinical management.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting the prognosis of cervical cancer patients with lymph node metastasis.
• To identify the most effective model for predicting survival outcomes.
• To support personalized treatment strategies.

Main Methods:

• Utilized data from the SEER*Stat database (2000-2020) for 1016 patients with cervical cancer and lymph node metastasis.
• Constructed and compared various machine learning models: XGBoost, random forest, SVM, ANN, and Cox proportional hazards model.
• Validated model performance using C-index, AUC, accuracy, and precision on training, testing, and an independent validation set.

Main Results:

• XGBoost demonstrated superior predictive performance with an AUC of 0.787 on the validation set.
• XGBoost achieved C-index values of 0.900 (training) and 0.773 (testing).
• Cox regression identified surgery at the primary site as a significant factor improving survival and reducing mortality.

Conclusions:

• The XGBoost model is highly effective for predicting prognosis in cervical cancer patients with lymph node metastasis.
• This machine learning approach offers valuable support for personalized clinical decision-making.
• Prognostic prediction can guide tailored treatment plans for improved patient outcomes.