CAL-Tutor: A HoloLens 2 Application for Training in Obstetric Sonography and User Motion Data Recording

Area of Science:

• Medical Education
• Medical Imaging
• Human-Computer Interaction

Background:

• Obstetric ultrasound (US) training is challenging, requiring trainees to correlate fetal anatomy with US images to achieve standardized views.
• Significant inter-operator variability exists in US examinations, highlighting the need for improved training methods.
• Current training methods often struggle to effectively teach the complex spatial relationships involved in fetal US navigation.

Purpose of the Study:

• To develop and evaluate the CAL-Tutor system, an augmented reality (AR) based training tool for obstetric ultrasound.
• To assess the potential of AR guidance to reduce the learning curve and improve spatial awareness for US trainees.
• To investigate the feasibility of capturing and analyzing user motion data for learning assessment.

Main Methods:

• The CAL-Tutor system integrates a US scanner, a physical phantom, and HoloLens 2 augmented reality headset.
• AR visualizations display the fetal model and corresponding US slices in the trainee's physical space.
• Data streams including eye gaze, head/hand position (HoloLens 2), US probe position (ARToolkit, NDI Aurora), and external camera footage were recorded.

Main Results:

• Demonstrated the system with engineers, showing it aids novice users in navigating the US probe to standard planes.
• Successful data capture of user motion and learning processes.
• Initial data visualizations indicate meaningful insights into user behavior can be extracted.
• Positive user feedback suggests improved assessment with AR guidance.

Conclusions:

• The CAL-Tutor system shows promise as an AR-based tool for obstetric ultrasound training.
• AR guidance appears to facilitate novice user navigation and improve spatial understanding.
• The system's data capture capabilities offer a rich resource for analyzing user performance and learning trajectories.