Visual Augmentations for Ultrasound Assessment Training of Medical Students

Research Background and Problem

• Problem and Challenges: The paper highlights the challenges faced by medical students in using ultrasound technology for assessments in emergency medical situations (e.g., E-FAST, focused on trauma evaluation), particularly in acquiring high-quality images and understanding how parameters affect imaging. Traditional training methods rely on individual guidance from experts, which is both resource-intensive and costly.
• Significance: E-FAST is critical for evaluating conditions such as internal bleeding, providing key diagnostic information in a rapid and non-invasive manner. Enhancing students' understanding of ultrasound imaging can not only improve learning outcomes but also increase accuracy and efficiency in future practical applications.
• Research Motivation and Related Work: While technologies like Virtual Reality (VR) and Augmented Reality (AR) have seen some application in medical training, they still have limitations (e.g., lack of realistic tactile feedback, reliance on fixed scenarios). In contrast, this paper explores a novel approach using visual augmentation technology to support student learning, aiming to bridge the gap between traditional manual education and virtual simulation training.

Solution

• Method and Solution: The authors designed four different visual augmentation methods (Location, Iconography, UI Indication, Amplification) to help students better understand and adjust two key ultrasound image parameters: Depth and Gain.
• Innovation: Unlike fully automated AI diagnostic support tools, these augmentation designs aim to encourage students to reflect on the learning process rather than blindly follow system prompts. This non-intrusive support approach seeks to maintain the ecological validity of real-world training scenarios.
• Implementation Steps:
  1. Needs Analysis: Participated in ultrasound courses, studied students' learning processes, and discussed with experts the main challenges students face in ultrasound imaging.
  2. Augmentation Mode Design: Based on observations, four augmentation modes were designed and implemented:
     • Location: Added arrows on the interface to indicate the problem area.
     • Iconography: Overlaid icons on the ultrasound image to show the type of issue.
     • UI Indication: Highlighted the depth or gain indicators on the interface.
     • Amplification: Visually exaggerated image quality issues caused by incorrect parameter settings.
  3. Study Design and Evaluation: Used the Wizard of Oz method (simulated automation) to ensure experiments closely resembled real-world scenarios with clinical equipment and simulated patients.
  4. Data Collection and Analysis: Employed quantitative Likert-scale questionnaires and qualitative interviews to analyze the impact of augmentation methods on user confidence, understanding, confusion, and distraction effects.

Research Outcomes

• Specific Findings:
  • The Location augmentation mode was considered the most helpful in enhancing confidence and understanding while reducing confusion and distraction.
  • The Amplification augmentation mode encouraged students to reflect more deeply on the reasons and consequences of parameter adjustments, supporting deeper learning, though it was barely noticed by some students.
  • The Iconography augmentation mode performed poorly due to obstructing the image and affecting students' attention allocation.
• Advantages Compared to Existing Solutions:
  • Compared to traditional teaching methods, these visual augmentation methods are more personalized and do not require full-time expert supervision.
  • Unlike fully AI-based tools, these augmentations support learning while avoiding complete reliance on technological guidance.
• Experimental Results:
  • Quantitative analysis showed varying perceived support effects among participants for different augmentation modes, with the ‘Location’ mode performing the best.
  • Qualitative analysis revealed that students' preferences for augmentation modes depended on the salience of the augmentation and the opportunities it provided for reflection.
• Limitations and Future Directions:
  • Since the experiment relied on the Wizard of Oz method, augmentations were manually triggered by researchers. Future work should explore real-time automatic detection to trigger augmentations.
  • The current study used a small-scale, single-location sample, limiting the generalizability of the results. Future research should expand to diverse medical scenarios and larger-scale testing.
  • The evaluation focused primarily on parameter adjustments and did not explore how other critical operations (e.g., manual manipulation of the transducer) could be supported through visual augmentation.

Through this study, the authors not only demonstrated the potential of visual augmentation technology in supporting medical students' learning of ultrasound skills but also proposed various ways to enhance student reflection and independent learning. This provides new perspectives and practical directions for the design and application of technology in medical training.