Accuracy is essential for optical head-tracking in cranial radiotherapy. Recently, the exploitation of local patterns of tissue information was proposed to achieve a more robust registration. Here, we validate a ground truth for this information obtained from high resolution MRI scans. In five subjects we compared the segmentation accuracy of a semi-automatic algorithm with five human experts. While the algorithm segments the skin and bone surface with an average accuracy of less than 0.1 mm and 0.2 mm, respectively, the mean error on the tissue thickness was 0.17 mm. We conclude that this accuracy is a reasonable basis for extracting reliable cutaneous structures to support surface registration.
, Sohma Y, Saito T, Kin T, Oyama H, et al. Marker-lesstracking of brain surface deformations by non-rigid registration integrating surface and vessel/sulci features. Int J Comput Assist Radiol Surg 2016;11:1687–701. 10.1007/s11548-016-1358-7 26945999 Jiang J Nakajima Y Sohma Y Saito T Kin T Oyama H Marker-lesstracking of brain surface deformations by non-rigid registration integrating surface and vessel/sulci features Int J Comput Assist Radiol Surg 2016 11 1687 701  Horn BK, Schunck BG. Determining optical flow. Artif Intell 1981;17:185–203. 10
target maintenance task. Photos and pictures can also be useful to set up a feature-based tracking for the machine and its main parts. This means that the machine and its components can be recognized on the basis of their shapes or other key features, avoiding the use of markers (marker-lesstracking systems). Otherwise, AR systems typically make use of markers to identify machines, equipment or safety devices (marker-based tracking systems). In this case, however, a marker-less solution was preferred, as placing markers on the machine or on elements/parts of the
100 Hz. Current systems, such as Ascension MotionStar, can operate wired or wireless
and provide up to 20 sensor positions per target. The accuracy of these systems depends
on the structure of the magnetic field and is typically better, the closer the sensor is to
the field generating unit.
860 V. Methods
2.2. Marker-lesstracking systems
Marker-lesstracking systems only require the sensor devices and no artificial enhance-
ments of the environment or the target. They also come as inside-out or outside-in sys-
tems. Examples for inside-out systems are
such as the systems from ART (ART, 2011) and Vicon (Vicon Motion Sys-
tems, 1984). ‘Inside-out’ systems work the other way round in that the
markers are used as landmarks in the environment and the tracking devices
are attached to the relevant body parts. A well-known example for inside-
Framing Multimodal Technical Communication 611
out tracking is the Wii-Remote developed by Nintendo. Several technol-
ogies can be used for tracking, such as visual tracking in different domains
(for example infra-red) and electromagnetic tracking.
2. ‘Marker-lessTracking Systems