Expert variability as a benchmark for validating automatic CT–MRI registration in brain stereotactic radiosurgery

Key Points

• The research aims to quantify expert variability to establish benchmarks for validating CT-MRI registration in brain radiosurgery.
• Assessed multi-expert variability in CT-MRI registration.
• Developed a probabilistic framework for algorithm validation.
• Evaluated an AI-driven convolutional neural network (CB) algorithm in practical scenarios.
• Established clinically relevant thresholds for CT-MRI registration accuracy.
• Demonstrated high reliability of the AI-driven algorithm across scenarios.
• Showed reduced operator dependence in CT-MRI registration processes.