Progress in Development of Lung Cancer Survival Prediction Models Using Machine Learning Based on SEER Database

Area of Science:

• Oncology
• Data Science
• Biostatistics

Background:

• The Surveillance, Epidemiology, and End Results (SEER) database is a key resource for cancer data.
• Machine learning (ML) is increasingly utilized for developing clinical prediction models.

Purpose of the Study:

• To review the application of ML algorithms in lung cancer survival prediction models (LCSPMs).
• To identify challenges and future directions in ML-based LCSPMs.

Main Methods:

• Review of ML algorithms: logistic regression (LR), support vector machines (SVM), decision trees (DT), random forest (RF), artificial neural networks (ANN), and extreme gradient boosting (XGBoost).
• Analysis of their use in constructing LCSPMs from the SEER database.

Main Results:

• Various ML algorithms have been applied to develop LCSPMs.
• Identified challenges include data imbalance, poor model interpretability, and insufficient external validation.

Conclusions:

• ML shows promise for improving lung cancer survival prediction.
• Future research should focus on addressing current limitations for more robust and reliable models.