The Factory Times is the Student-Run school newspaper for SUNY Poly.

Five Ways A.I. Is Making Us Healthier

Five Ways A.I. Is Making Us Healthier

In the super-hyped world of artificial intelligence, it's easy to forget or even overlook the ways computer systems can help keep us healthy, and not just by keeping us smart. Right now, A.I. is working behind the scenes and under the direction of scientists, doctors, and researchers around the world to bring us new pharmaceutical therapies, new ways to triage human illnesses, faster ways to diagnose them, even new ways to document treatment, and digital ways to keep physically fit. I’ll briefly touch on all five ways A.I. is making us healthier right now in 2020.

1. ) New Drug Discoveries

Researchers have managed to train an A.I. system with enough base knowledge to help pick out useful chemical compounds for drug development and manufacturing processes. Unlike other realms of science, duplicating the results of an A.I. system can be challenging, and thus the data produced by artificial intelligence systems can be scientifically dubious. Current implementations don’t do a very good job of showing their work for other scientists to verify. Rather than deal in chemical understanding, however, A.I. allows teams to characterize compounds and predict chemical reactions, such as how our bodies might metabolize drugs.

2.) Triage of Symptoms

With our world bracing for the official second wave of COVID-19 infections, and flu season about to be upon us as well, A.I. researchers are putting their computer brains to the test in the triage of the potential ‘twindemic.’ The ability to differentiate between COVID-like symptoms and flu-like symptoms is at the forefront of the administrators of hospitals and clinicians alike. A.I. is currently demonstrating the ability to digest data collected on symptoms, as well as diagnostic information from X-rays, MRIs, and CT scans, to aid in the rapid triage and separation of flu infections versus COVID-19 infections. The challenge, as is often the case with artificial intelligence, is how to train the brain on the most important indicators and filter out the rest of the noise without as much human intervention.

3.) Diagnostic Testing

Though not my favorite part of life, after triage in an emergency room, a patient is usually passed on for additional diagnostic testing to make an initial diagnosis. As previously mentioned, this can be in the form of diagnostic imaging such as CT scans, but it can also be laboratory or blood work. In keeping with the current events of the coronavirus, researchers at Oxford have trained an artificial intelligence to discern pieces of the virus in a given sample, bypassing the idea of sorting out antibodies all together. This not only reduces the amount of time needed to determine infection (a whole five minutes), but it also forgoes the need for samples to be prepared or mixed to get an accurate result.

4.) Better Clinical Notes

Ask a clinician or researcher, and they may tell you, their practice is only as good as their notes. When it comes to science in general, the notes we take while we’re designing, setting up, and executing an experiment are also key to our ability to reproduce results, write about them, and have others do the same. What artificial intelligence brings to the table, however, is the ability to almost digest often highly technical notes into a structured format that’s capable of generating additional data through its intrinsic statistical value. Determining that value and classifying the information contained in the notes is where we sit today. Companies now more than ever are trying to squeeze every last byte of value out of their data, while health care companies, clinics, doctors, and hospitals are no different. Artificial Intelligence offers a set of anonymous eyes that can scour millions of pages of medical records without ever looking at protected personally identifiable information. The benefits of pooling this data to sample from on such large scales are seemingly endless to patients, providers, and of course, researchers.

5.) Digital Fitness

We may have been headed in this direction before the pandemic, but seven months in, people are talking to their bedroom mirrors. Oh, wait, that’s not a mirror; it’s a home fitness training machine. That’s right, number five is the last thing I would’ve ever thought about in terms of A.I. making us healthier, digital fitness. The way A.I. works its way into our homes is often packaged inside of otherwise mundane machinery, and home gym systems are no longer exempt. In this case, however, artificial intelligence is all about stroking our ego and keeping us motivated. The A.I. components allow trainers to record a library of exercises,  then digitally slice and dice the footage in preparation to customize sessions across their entire user base. Trainers’ faces are scanned and voices are digitally mapped so the A.I. can make them appear to say any user’s name and deliver totally personalized experiences. Then, as a user signs-in to begin their routine, the machine splices together all the components of that day’s workout, cross referencing their personal preferences and stated fitness goals.


Conclusion

All of this doesn’t automatically make us any healthier, nor does it make us any smarter. What drives the value in all of this is the ways in which A.I. enriches our human experience. New drugs, rapid triage and testing for COVID-19, and better storage of clinical information are working for us right now in the United States. The engines that power machine learning are all around us now, so if 45 minutes in front of a fancy mirror or on a fancy bicycle is what you need to feel better in these uncertain times, then by golly, work it out.

Censorship (Part 2): To Block or Not To Block

Censorship (Part 2): To Block or Not To Block

The Transfer of Life

The Transfer of Life