Comparing the Effectiveness of Artificial Intelligence Models in Predicting Ovarian Cancer Survival: A Systematic...

Area of Science:

• Oncology
• Medical Informatics
• Data Science

Background:

• Ovarian cancer (OC) survival prediction is critical for patient management.
• Machine learning (ML) offers advanced analytical capabilities for prognostic modeling.
• Systematic review to assess ML algorithm effectiveness in OC survival prediction.

Purpose of the Study:

• Evaluate ML algorithms for predicting overall survival (OS), recurrence-free survival (RFS), progression-free survival (PFS), and treatment response in ovarian cancer.
• Identify key features influencing the predictive accuracy of ML models for OC.
• Assess the current landscape and future directions of ML in ovarian cancer prognostication.

Main Methods:

• Systematic literature search of PubMed, Scopus, Web of Science, and Cochrane databases.
• Inclusion of 32 studies published within the last decade, with a focus on recent advancements post-2021.
• Analysis of commonly used ML algorithms (e.g., random forest, SVM, deep learning) and evaluation metrics (AUC, C-index, accuracy).

Main Results:

• Common ML algorithms include random forest, support vector machines, logistic regression, XGBoost, and deep learning.
• Area Under the Curve (AUC), concordance index (C-index), and accuracy are frequently used evaluation metrics.
• Significant predictors identified include age at diagnosis, tumor stage, CA-125 levels, and treatment factors.

Conclusions:

• ML models show significant potential for predicting ovarian cancer survival outcomes.
• Challenges in model accuracy and interpretability need to be addressed.
• Integrating diverse data types (clinical, imaging, molecular) with multimodal ML approaches is crucial for enhanced prognostic precision.