Evaluating the Utility and Limitations of Machine Learning Tumor Segmentation for Automated Longitudinal RANO Treatment Response Classification

Key Points

• The research aims to assess the effectiveness and challenges of using automated tumor segmentation in treatment response evaluation.
• Evaluated machine learning models for tumor segmentation in treatment response assessments across multiple timepoints.
• Analyzed the impact of post-surgical enhancement on segmentation accuracy and treatment response classification.
• Automated segmentation models showed limitations in accurately assessing treatment responses due to nuances in tumor appearance over time.
• Notably, changes in tumor location and morphology between different timepoints affected classification accuracy.