FocuGazeDx: Real-Time Eye-Tracking Framework forRadiological Image Analysis

The examination of ocular movements, known as gaze analysis, has a wide range of applications due to the current development of sensor technology, which can now collect biometric data. Gaze analysis has succeeded in Human-Computer Interactions (HCI) in recent years. Eye Gaze Trackers (EGTs) are devices that can estimate the direction of a person's gaze, monitoring and recording the movement of the eyes. This work focuses on using eye trackers in the healthcare sector and how they can support doctors, especially radiologists, during their analyses and diagnoses. The proposed framework, FocuGazeDx, uses eye-tracking technology to monitor and analyse doctors' visual attention while they analyse medical images, with the aim of reducing diagnostic errors caused by distraction or fatigue. The final result is the detection of unattended or under-explored regions and the generation of post-observation visual feedback for the doctor. The main components of the proposed architecture are: Real-Time Eye-Tracking Data Acquisition, Data Fusion Module, Coverage Analysis Engine, Reinforcement Learning Module, and Visual Feedback Interface. This framework aims to overcome limitations of previous approaches by collecting fixation data and introducing a predictive and adaptive component, moving towards precision, personalized, and intelligent diagnostic support systems.