Researchers at Macquarie University’s Computational NeuroSurgery (CNS) Lab have developed a novel framework for quantifying visual expertise by analyzing eye-movement patterns through fractal mathematics. Led by Professor Antonio Di Ieva, the team’s findings demonstrate that the transition from novice to expert in medical imaging is marked by measurable structural changes in how individuals scan visual data. These insights now provide a scalable method for training artificial intelligence to interpret complex medical images with expert-level efficiency. The research builds on the premise that visual expertise extends beyond accumulated knowledge to include organized, non-random scanning behaviors. To test this, the CNS Lab conducted two independent studies tracking the eye movements of medical students, surgical trainees, and qualified neurosurgeons as they examined X-rays, CT scans, and MRI images. By applying fractal analysis to high-resolution gaze data, researchers translated the complexity and regularity of visual scanning into numerical expressions. Dr. Ghasem Azemi, who co-authored the work published in Medical & Biological Engineering & Computing, explained that fractal metrics capture the shift from low-level feature detection to targeted, diagnosis-driven processing. A longitudinal study of thirteen medical students revealed a statistically significant progression in the Fractal Eye-Gaze Expertise Index (FEI). As students advanced through their curriculum, their gaze patterns became less erratic and more structurally aligned with those of seasoned specialists. A parallel study published in the Journal of Eye Movement Research, involving sixty-nine participants, expanded these findings. The data showed that expert neurosurgeons consistently allocated more fixation time to pathological regions and processed image complexity differently based on the specific clinical presentation. Leveraging these behavioral signatures, the team trained a machine-learning model that integrated gaze duration and three-dimensional fractal dimensions. The algorithm successfully distinguished between naive observers, trainees, and qualified neurosurgeons with an accuracy exceeding ninety-three percent. According to Professor Di Ieva, the project originated from a core objective in computer vision: translating human cognitive shortcuts into reproducible data streams for machine training. The research confirms that expertise fundamentally reorganizes the architecture of visual attention, leaving a quantifiable fractal signature across all stages of professional development. The implications for both medical education and artificial intelligence are significant. The FEI framework offers an objective, scalable metric for evaluating clinical trainee progression, potentially refining neurosurgery and radiology curricula. Simultaneously, the gaze-derived data provides a robust training dataset for AI systems designed to automate medical image analysis. By encoding expert scanning patterns into algorithmic learning, the CNS Lab has established a foundation for diagnostic AI that mimics human expertise rather than merely processing raw pixel data. This convergence of cognitive science, fractal geometry, and machine learning positions the framework as a critical tool for the next generation of automated clinical diagnostics.