IEEE Computational Intelligence Magazine - May 2023 - 11

Gary B. Fogel
Natural Selection, Inc., USA
Industrialand
Governmental
Activities
Piero P. Bonissone
Piero P. Bonissone Analytics, USA
AI in Healthcare and Life Science
I
n a previous communication [1],
we highlighted the inaugural 2023
IEEE Conference on AI (IEEE
CAI 2023) - https://cai.ieee.org/2023/
- our first conference with an industry
focus. IEEE CAI 2023 is structured
along six verticals, covering Industrial
AI, AI in Healthcare/Life Science, Transportation/Aerospace,
Energy, Earth System
Decision Support, and Social Implications of
AI/Privacy (Fig. 1). Here we focus on
the second vertical, AI in Healthcare and
Life Science.
Healthcare is itself a very large field,
covering everything from the development
of novel therapeutics to medical
imaging to wearable sensors to detecting
incipient medical issues to in-home monitoring.
Life science is its own additional
large field focusing on the understanding
of genomics, proteomics, systems biology,
and the mapping of genotypes to
phenotypes in light ofenvironmental and
developmental dynamics. The scientific
lessons learned in life sciences are transformed
into improved healthcare outcomes
on a daily basis [2].How does
AI/CI play a role?
Even just within the process of drug
development, there are many areas where
AI plays an active role. A new biological
target of interest can be identified using
AI to help screen the literature or other
datasets. This target, typically a protein,
Digital Object Identifier 10.1109/MCI.2023.3248699
Date ofcurrent version: 6April 2023
may play an important role in a biochemical
pathway. The understanding of that
role in the entire system ofprotein interactions
in a cell can be enhanced through
AI by studying protein-protein interactions
and inferring novel connections. A
pharmaceutical company may be interested
in developing small molecule compounds
that bind to and block the active
site of this target protein. A search begins
in silico for small molecule candidates that
bind to the target. Many of these tools
make use of aspects of computational
intelligence to search large spaces ofsmall
molecule chemistries to identify putative
novel drug candidates. The topmolecules
from this process are then considered candidates
for lead optimization to refine
them for their suitability as a drug as the
compoundmust not only have the proper
inhibitory function but have the proper
solubility and transport to get to the locations
where its activity have the intended
effect over time. AI can be used to help
medicinal chemists make educated
changes to small molecules to improve
their efficacy. Each drug candidate (now
termed as a medical entity (NME)) then
enters a process ofphased clinical trials of
increasing cost and stringency. AI can
assist in predicting the success or failure of
these candidates in this pipeline or assist in
the repurposing ofNMEs should they fail
at some point during the process. Given
Corresponding author: Gary B. Fogel (e-mail:
gfogel@natural-selection.com).
the clinical trials are costly, AI can also be
used to optimize the trials themselves or
help understand other barriers to
approval. Should theNMEpass all clinical
phases and be approved, it can then be
cleared for market delivery to improve
human health. Given the average cost of
this entire process from start to finish is
roughly US$1 billion per NME that succeeds,
any advantages that AI can provide
in accelerating the process or improving
the odds ofsuccess are valued highly.
However, the applications of AI do
not stop there. Each drug that has been
brought to market now has to be produced
efficiently at scales and then distributed
nationally or globally. AI can
assist with supply chain optimization or
prediction of market demand. Further,
medical personnel needs to be educated
on when to use the new drug, and
patients made need to be monitored to
ensure that they are adhering properly to
their drug regimen. In these scenarios,
AI can also provide assistance; however,
one must be mindful to remain sensitive
to ethics and a requirement for patient
approval prior to 24/7 monitoring.
Patients under medical care are reviewed
regularly, and in some cases AI can be
used to alert physicians to deviations that
might have adverse effects.
These and other similar challenges
will be addressed at the IEEE CAI 2023
conference in the AI in Healthcare vertical.
How do we identify companion
diagnostics that look at the genetic
MAY 2023 | IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE 11
https://cai.ieee.org/2023/
IEEE Computational Intelligence Magazine - May 2023

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - May 2023
