Artificial intelligence-assisted machine learning models for predicting lung cancer survival.

Area of Science:

• Oncology
• Medical Informatics
• Data Science

Background:

• Lung cancer survival prediction is crucial for patient management and treatment planning.
• Machine learning offers advanced analytical capabilities for complex health data.
• Integrating AI tools like large language models (LLMs) can enhance predictive modeling in healthcare.

Purpose of the Study:

• To assess the feasibility of using large language model-Advanced Data Analysis (LLM-ADA) for developing machine learning models to predict lung cancer patient survival.
• To explore the implications of LLM-ADA in predicting survival outcomes for nursing practice.
• To identify key predictive factors for lung cancer survival using AI-driven analysis.

Main Methods:

• Retrospective study design utilizing a lung cancer patient dataset.
• LLM-ADA employed to construct and evaluate three distinct machine learning models.
• Model performance assessed using calibration plots for reliability.

Main Results:

• The study included 737 lung cancer patients with a 73.3% survival rate.
• Calibration plots confirmed robust reliability across all developed models.
• The Random Forest model achieved the highest predictive accuracy, with key features including preoperative white blood cells and lung function (FEV1).

Conclusions:

• LLM-ADA effectively supports the creation of machine learning models for lung cancer survival prediction.
• The study highlights the significant role of preoperative factors in predicting patient outcomes.
• Findings suggest LLM-ADA empowers non-technical healthcare professionals in utilizing advanced analytics for clinical decision-making.