AI to be used to fight against opioid addiction
The fight over opioid addiction is unlike any health crisis in recent history. Access to medication designed only for prescription use can sometimes be as easy as clicking a link on an Instagram ad. When doctors overprescribe them, the problem is only made worse.
In 2015, there were on average 91 deaths every day due to opioid overdose according to a White House report. Five years later, that number has grown by over 40 percent despite nationwide efforts to curb the opioid addiction epidemic.
As new techniques and innovations are made in medicine, the potential for drug abuse grows. Opioids such as fentanyl, a painkiller 50 times more potent than heroin, have become highly sought-after for their strong, immediate effects. A single dose of over 2 milligrams is enough to be lethal.
Jonathan Danfifer, sophomore chemistry major at California Baptist University, said the crisis is concerning.
“When you hear about those kinds of numbers, it makes you wonder how we can actually solve a national problem like this,” Danfifer said.
Even after attempted reforms in the way medications are prescribed to patients and a declaration of a national health emergency to increase funding, optimism among medical experts is scarce.
Due to ease of access, it is more likely for someone to be hooked on opioids than many illegal drugs.
“It’s a big problem,” said Dr. Jenifer Nalbandian, professor of chemistry.
“Opioids change the chemistry within the brain. They can be dangerously addictive.”
One of Nalbandian’s courses at CBU even discusses the chemistry of opioids. The painkillers most commonly abused are difficult to track, as about two out of five of them come from family members who have leftover medicines from prior prescriptions.
In Riverside County, deaths from opioid overdoses are 10 percent higher than the current national average according to a report by the Riverside University Health System. To fight this, the county was awarded a $7 million grant to combat the crisis locally.
One way officials are looking to do that is through better information services — asking questions such as who is more likely to overdose, where is the problem most concentrated and how long do people stay addicted. That approach is becoming commonplace as the fight against addiction looks for more creative solutions.
Innovations in the field of artificial intelligence have sparked interest in what can be done to fight addiction with computing power. Researchers are taking a multifaceted approach with AI; some are looking at ways to curb the sale of illegal opioids online, while others are looking to get datasets for prevention and outreach.
A federal Health and Human Services report indicates that a comprehensive structuring of machine learning on social media will help siphon out illegal sales through sites such as Facebook and Instagram, where sellers use clever tactics to get around web restrictions, even using deliberate misspellings or a system of verification links to make their operations difficult to track.
Training algorithms to detect illegal opioid operations would be an evolving task, as researchers adapt to new tactics used to avoid their searches.
Mark Kim, assistant professor of computing, software and data sciences, said algorithms provide a good sweep of the big data seen on social media sites.
“You can’t test against every case,” Kim said. “But the data is increasing, so you have to rely on technology.”
That means sorting through the millions of daily posts on social media sites, with human-trained algorithms intelligently sorting through the data it finds to be potentially malicious. That is where humans, and in this case, law enforcement, come in to make the final call. As the programs sort through more data and gain experience, they become more reliable, but they are never foolproof.
While artificial intelligence admittedly will not solve the epidemic by itself, researchers and public health officials indicate it will represent a big step in the right direction toward helping Americans find independence from opioid addiction.