Summary and Future Trends - Computational Intelligence in Biomedical Engineering - page 329

CI in biomedical
engineering

Disease prognosis prediction

Diagnosis of disease
•
•

Rehabilitation-rate
monitoring
New and no
investigation yet.
Will require regression
modeling using CI
techniques.
Direct use of
regression modeling
possible.

Classification.
20+ years of
literature.
Includes NNs, SVM,
fuzzy logic, HMM, etc.
Research in better
feature extraction for
more accurate
diagnosis.

• New with little to no results

using CI techniques.
Will require time-series
analysis and prediction
techniques.
Extensions of current CI
possible, e.g., time-series
prediction using SVMs.

•

•

•

•

•

•

•

FIGURE 8.1
Some of the current and future research areas for CI techniques in biomedical
engineering.

for example, dosages could be predicted and pharmacies better prepared for
the demand.
CI prediction could also be used, as an early warning system, for exam-
ple, using a time-series analysis of MTC variables, it may soon be possible
to predict the risk of falling (see Section 7.4.1.3). Research in this direction
is relatively new but a successful outcome requires a suitable application of
CI technology. In rehabilitation-rate monitoring a similar prediction objective
might be pursued, the idea here is to see how quickly a patient can regain
normal function. This will require regression CI techniques because we are
monitoring the recovery process, but similar problems are present regarding
accuracy, index measures, and the quantity of data. Both prognosis prediction
and rehabilitation-rate monitoring are important applications in the treat-
ment of a patient recently diagnosed with a debilitating disease. These phases
will benefit considerably from CI techniques, which have now been introduced
to medicine.
A further emerging trend in healthcare is the development of automated
healthcare systems also known as healthcare information technology (IT).
Some of these exciting new technologies will be considered in the following
sections and there is a broad avenue of possibilities for the implementation of
CI techniques in the healthcare IT. In addition to research into the intercon-
nectivity of more advanced monitoring systems, there is a need to examine the
monitoring devices themselves. There are now many portable devices that can
monitor the body's vital signs such as portable ECG machines. More research
is required to miniaturize these devices and provide them with wireless capa-
bilities, so that monitoring can be done not only in hospitals but also outdoors
and at home. These possibilities introduce further research questions, ranging
from the eciency of the communication networks, the use of biomedical