Evaluation of Areal Touch Feedback for Palpation Simulation

Abstract

The effectiveness of the areal contact versus the point contact was experimented. We created virtual 3D cubic volumetric objects consisting of approximately 500,000 nodes. The object is placed on a plane so that the nodes at the bottom of the object are constrained. A user can interact with the object by pushing and pulling at the top surface of the object as shown in Figure 2. Figure 2 (a) and (b) are the configurations of pulling up, and pushing down, respectively, in the middle of the top surface with the haptic interface. The square-shaped areal contact was made with the tactile display unit attached to the gimbals of Phantom TM haptic device. For a point-based contact, only the Phantom haptic device was used for interaction with the virtual object. We constructed two volumetric soft objects where four hard blocks, which represents tumors, are placed inside of each soft volume as illustrated in Figure 3. Figure 4 shows the top view of the test volume objects. The test object (a) is used for the experiment with the point haptic feedback only, while the test object (b) is used for the experiment with the area-based haptic feedback with the augmented tactile display. We asked 20 human subjects to explore the objects with the haptic interface device, one without the tactile display unit and the other with the tactile display unit. Their task was to locate the hard portions (i.e., tumors) inside the volume, and they were asked to draw the tumors they found on a piece of paper. Each subject drew what he/she visualized the tumor’s location and size solely with the touch feedback. All subjects did not have a pre-knowledge of the number of tumors that they can find. Figure 5 shows the representative drawings done by the subjects. We can observe that the area-based haptic interface gave a superior results compare to the one with the point-based interface. With the point-based interface, most subjects missed the tumor #4 which is relatively small in size. However, all of them were detected with the area-based haptic interface. In palpation, it is important to find not only the number of tumors but also a precise location and the size of the tumors. We defined the accuracy measures concerning the center location of the tumor and the actual size of the tumor. Table 2 and 3 shows the average errors computed against the accuracy measures. Although there seems to be an intrinsic source of error due to human perception, our results clearly demonstrates that areal haptic feedback provides a better visualization of the object.