Biomedical Engineering Reference
In-Depth Information
628114
645976
6
90434
FIGURE 1.53
Supervised learning using RadViz for the three-class classification of cancer cell lines into melanoma (squares),
leukemia (circles), and other or nonmelanoma, nonleukemia (triangles) classes. The RadViz algorithm classified
the known cancer cell lines using a 1400-compound subset of the NCI database. The compounds' GI50 values
were used to carry out the classification by selecting those compounds best at classification at the
p
< 0.01 statis-
tical criterion with Bonnferroni correction for multiple tests. The compounds whose GI50 values were most effec-
tive and met the statistical criterion for classification have their NSC numbers shown in the class sector. In the
RadViz algorithm, the positions of these compounds and the GI50 values are what pulls each of the 60 cells
(points) into their respective positions, forming a visual output of the supervised clustering algorithm. Reprinted
from McCarthy, J.F., Marx, K.A., Hoffman, P.E., Gee, A.G., O'Neil, P., Ujwal, M.L., Hotchkiss, J. (2004).
Applications of Machine Learning and High-Dimensional Visualization in Cancer Detection, Diagnosis and
Mangement. In: Umar, A., Kapetanovic, I., Khan, J., eds.
The Applications of Bioinformatics in Cancer Detection, Ann.
N.Y. Acad. Sci.
1020:239-262.
biosensor data relevant to drug discovery, the process is general and potentially applies to
all biosensors possessing complex multidimensional data inputs.
1.5
Future Prospects for Biosensors
Looking to the future, there are many ways in which improvements can be made in creat-
ing smart biosensors. First, progress is continually being made in improving biosensor plat-
forms. For example, significant efforts are underway to miniaturize individual biosensor
components using microfabrication strategies and integrating them into MEMS devices
(182). As a result, in a single MEMS device samples can be preprocessed and, in principle,
be subsequently analyzed by multiple biosensor tests. In many cases, the miniaturization
