Information Technology Reference
In-Depth Information
Total error rate (the proportion of error instances to all instances):
FN
+
FP
TER
=
=
23
.
FP
+
TP
+
FN
+
TN
Semi automatic detection gives far better results then full automatic. 77.0% of
tested maps were rightly classified and the visible lesions were detected and de-
scribed identically to radiologist diagnosis. The FPR error appears more often then
FNR. This situation happens because its more complicated task to estimate lesion
presence and its position then only reject the hypothesis of presence of lesion.
Most of the errors were caused - similarly to full automatic detection case - by
overestimation of asymmetry regions (i.e. additional regions between brain he-
mispheres or in the top / bottom of the perfusion map).
The errors might be eliminated by more accurate detection algorithm with addi-
tional adaptive factors i.e. taking into account the brain volume (for adaptation the
minimal lesion size factor), the noise ration in perfusion map (adaptive median
filter size) or average perfusion in whole hemisphere (adaptive threshold for lesion
acceptance).
7 Data Visualization - Augmented Reality Environment
Augmented reality (AR) is a technology that allows the real time fusion of com-
puter generated digital content with the real world. Unlike virtual reality (VR),
that completely immerses users inside a synthetic environment, augmented reality
allows the user to see three dimensional virtual objects superimposed upon the real
word [11].
Analyzing the articles and conference proceedings of the leading AR research
symposium several significant research directions on this field can be identified;
those are: tracking techniques, display technologies, mobile augmented reality and
interaction techniques.
Tracking techniques is a group of computer methods that are used to achieve
robust and accurate overlay of virtual imagery. There are various optical tracking
techniques including fitting a projection of a 3D model onto detected features in
the video image, matching a video frame with photos and real time image registra-
tion algorithms [35]. One of the most popular methods is a calculation of the 3D
pose of a camera relative to a single square tracking marker. The task performed
by tracking software is detection of presence of tracking marker and determination
of its position based on interpretation of special pattern that is printed on it. The
commonly used display technologies are head mounted, handheld and projection
displays for AR. Many contemporary mobile devices like iPods and cell phones
can execute sophisticated algorithms and are equipped with high resolution color
displays and cameras. For those devices AR applications for outdoor usage are
developed. The last type of methods is interaction techniques that enable user inte-
raction with AR content.
Augmented reality shows its usefulness especially in the field of the medicine
[37]. The most notable examples are deformable body atlases, AR surgical navigation
Search WWH ::
Custom Search