Exploring interaction paradigms for segmenting medical images in virtual reality.

Area of Science:

• Medical imaging
• Virtual reality (VR)
• Human-computer interaction

Background:

• Virtual reality (VR) offers immersive platforms for medical image segmentation, aiding anatomical understanding for training, diagnosis, and surgical planning.
• User interaction is crucial for manipulating and interpreting medical data within VR environments.
• Optimal interaction schemes and input devices for VR segmentation tasks are not yet clearly defined.

Purpose of the Study:

• To compare user performance and experience with two distinct input schemes for medical image segmentation in virtual reality.
• To evaluate the effectiveness of keyboard and mouse (KBM) versus motion controllers (MCs) for VR-based segmentation tasks.

Main Methods:

• Twelve participants performed segmentation on 6 CT/MRI images using both KBM and MC input methods.
• Performance metrics included accuracy, completion time, and efficiency.
• User-perceived performance and experience were assessed via a post-task questionnaire.

Main Results:

• No significant overall time differences were found between KBM and MC input methods.
• Keyboard and mouse (KBM) demonstrated faster segmentation for larger datasets.
• Accuracy remained consistent across both input schemes.
• Participants reported similar perceived challenge and efficiency for both methods.
• Motion controllers (MCs) were rated as more enjoyable to use.

Conclusions:

• Virtual reality segmentation software should offer adaptable input options based on task complexity.
• Enhancing motion controller interfaces could further improve usability and user experience in VR medical imaging.