Characterizing Visual Neurosurgical Expertise in Brain MRI Visualization Using Eye-Tracking and 3D Fractal Dimension...

Area of Science:

• Neuroimaging and neuroscience research.
• Medical image analysis and interpretation.
• Human visual perception and expertise.

Background:

• Eye-tracking studies visual behavior in medical imaging but lacks neurosurgeon representation.
• Understanding neurosurgeon visual expertise is crucial for training and computational model development.
• Expert visual search strategies in neurosurgery remain undercharacterized.

Purpose of the Study:

• To characterize neurosurgery-specific visual expertise using eye-tracking.
• To investigate how fixation duration and image complexity relate to expertise.
• To develop computational models for neurosurgical image interpretation.

Main Methods:

• Compared gaze behavior of naïve observers, registrars, and consultant neurosurgeons.
• Analyzed fixation duration and 3D fractal dimension (3DFD) of brain MR images.
• Utilized linear mixed-effects modeling and Random Forest classification.

Main Results:

• Experts showed longer fixation durations on pathological images.
• Expert visual search strategies were dependent on pathology type.
• Combined fixation duration and 3DFD improved classification accuracy (>93%) for pathology identification.

Conclusions:

• Identified neurosurgery-specific markers of visual expertise.
• Combined behavioral and image features can model expert strategies.
• Findings support the development of advanced training and computational tools for neurosurgical image interpretation.