A Novel Prediction Model for Colon Cancer Recurrence Using Auto-artificial Intelligence

Area of Science:

• Oncology
• Artificial Intelligence
• Biostatistics

Background:

• Colon cancer recurrence remains a significant concern for patients with stage II-III disease.
• Accurate prediction of recurrence is crucial for personalized treatment strategies.

Purpose of the Study:

• To develop and validate a novel artificial intelligence (AI) model for predicting recurrence in stage II-III colon cancer.
• To compare the predictive accuracy of the AI model against conventional statistical models.

Main Methods:

• A cohort of 787 patients with stage II-III colon cancer treated between 2000 and 2018 was analyzed.
• Binomial logistic regression was employed for conventional statistical analysis.
• An auto-AI software ('Prediction One') was utilized for recurrence prediction using the same dataset.
• Area Under the Receiver Operating Characteristic Curve (AUC) was used to assess predictive accuracy.

Main Results:

• The AI model achieved an AUC of 0.815.
• The conventional multivariate model had an AUC of 0.719.
• The AI model demonstrated a statistically significant improvement in predictive accuracy over the conventional model.

Conclusions:

• The developed auto-AI model offers superior accuracy in predicting colon cancer recurrence compared to traditional statistical methods.
• This AI model is accessible for development by clinical surgeons without prior AI expertise.
• The findings suggest a potential for AI to enhance clinical decision-making in colon cancer management.