Biomedical Engineering Reference
In-Depth Information
y
x
z
C
1
(Benign sample)
C
2
(Malignant sample)
Misclassified sample
5.0
2.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
6.0
3.0
3.0
5.0
3.0
10.0
3.0
5.0
3.0
9.0
10.0
10.0
10.0
10.0
5.0
10.0
10.0
10.0
FIGURE 5.12
(See color insert following page 330)
Two views of the same 3D colored glyph used to cluster 683 nine-dimen-
sional data vectors into two classes (benign and malignant). Vectors assigned to the benign class (
C
1
) are repre-
sented by the yellow facets, while those associated with the malignant class (
C
2
) are shown as red. Visual
identification of a misclassified data vector based on the color cue is shown in the right view.
epithelial cell size in the feature vector appear to have resulted in classifying it as a malig-
nant-type tissue. By maintaining a record of the inputs assigned to every cluster unit, it is
possible to trace those that contribute to significant changes in surface characteristics,
either shape or color, in the graphical form thereby indicating the presence of distinct
attributes as compared to the rest of the input vectors in the dataset.
5.4
Conclusions
Smart and intelligent biosensors utilize highly selective biological recognition elements to
accurately and rapidly identify specific stimuli, analytes, or biochemical reactions.
Although the concept of a biosensor has been around over several decades, it is only
recently with the introduction of new biomaterials, microfabrication technologies, and
advanced computing algorithms that it became possible to incorporate some semblance of
intelligence into the device. This chapter briefly examined a variety of key issues associ-
ated with biosensor design and presented an overview as to how AI and pattern recogni-
tion algorithms can enhance system performance and assist the analyst with interpreting
the acquired data (Table 5.2). In this regard, several adaptive connectionist computing
