Traditional Cox regression outperforms large language models in predicting long-term progression of intermediate to advanced hepatocellular carcinoma

Puntos clave

• The traditional Cox regression model predicts long-term progression of hepatocellular carcinoma more effectively than large language models.
• Significantly, Cox regression demonstrates superior performance when compared to LLMs in predicting risk, indicating a key advantage for established methods.
• Assessment using a combination of ablation and immuno-checkpoint inhibitors with standard treatment may further enhance progression-free survival.
• These findings highlight the need for integrating traditional models with new data processing methods for improved prediction accuracy.