Abstract

This work presents the design and evaluation of a gesture-controlled system for interactive manipulation of radiological images and 3D models using the Kinect device. Several abstractions have been implemented and refactored to improve the system performance, making the application simpler, at an affordable cost. Additionally, specific gestures to change the visualization settings of 3D models represented by layers were also successfully modeled. Further, we have conducted systematic and detailed usability testings with users to determine quantitative performance measures and qualitative analysis (usefulness, visual quality of the interface, ease of learning, ease of use, 3D spatial perception, level of interactivity, mental and physical fatigue, effectiveness and satisfaction). The results show that the participants are able to perform the tasks of search, selection and manipulation of 2D images (zoom in/out and translations) and 3D models (zoom in/out and rotations), quickly and accurately, demonstrating the usefulness of the system as a possible effective and competitive alternative solution, to the traditional use of the negatoscope.

Highlights

  • Interactive gesture controlled systems have gained nowadays greater visibility

  • We have used in our evaluation study a small number of user reviews (16), if there were any usability issues, they would have been highlighted

  • In our current evaluation study, the different users were behaving in a very similar way and no major usability error was found

Read more

Summary

INTRODUCTION

Interactive gesture controlled systems have gained nowadays greater visibility. This was essentially motivated by the evolution of gesture recognition devices, along with the improvement of techniques in Computer Vision. The gesture identification begins by tracking a certain segment of the human body skeleton, followed by the processing and treatment of the data obtained by tracking This procedure is usually performed by Support Vector Machines (SVM) [1], together with the training of neural networks to identify gestural patterns [7], [9], [17], [33], [3]. Considering the fact that these medical images are queried, searched and exchanged several times in the negatoscope (by the doctor or an assistant) to assist the surgeon in the decision making process during surgery, the asepsis in the operating room can be broken due to direct and accidental contact of the sterile gloves with the negatoscope or with the computer mouse/keyboard, putting at serious risk the patient’s life. There are times where there may be no other member of staff available, being the sole responsibility of the surgeon to access the images, necessarily requiring the exchange of gloves, followed by a sterilization procedure for cleaning the hands and arms (washing them with sufficient antimicrobial soap to ensure asepsis), after each use of the negatoscope

RELATED WORK
System Architecture
USER-BASED EVALUATION
RESULTS AND DISCUSSION
CONCLUSIONS AND FUTURE WORK
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call