New AI System Could Predict a Drug's Negative Side Effects

Researchers have developed an algorithm that can warn doctors about the negative side effects of drug-drug interaction.

Americans are taking more prescription drugs than ever. According to a survey, 55 percent of Americans regularly take a prescription drug.

What’s more, those who use prescription medicine take an average of four other drugs – from over-the-counter pills to vitamins and supplements. As the medication increases, so does the likelihood of interaction between those drugs.

Expectedly, this can trigger negative side effects, which range from long-term organ damage to even death.

Now researchers at Penn State have developed a machine learning system to address this issue. Basically, the AI system warns health professionals and their patients on the negative side effect that might occur when they mix different drugs.

In a statement, the Allen E. Pearce and Allen M. Pearce Professor of Industrial Engineering, Penn State, Soundar Kumara said:

“Let’s say I’m taking a popular over-the-counter pain reliever, and then I’m put on blood pressure medicine, and these medications have an interaction with each other that, in turn, affects my liver. Essentially, what we have done in this study is to collect all of the data on all the diseases related to the liver and see what drugs interact with each other to affect the liver.”

So, how does the AI system work?

Using AI to Determine the Negative Side Effects of Drug Interaction

Labeling thousands of drugs and millions of different combinations of possible interactions is time-consuming.

So, the researchers created an alert system using an autoencoder system – a form of artificial neural network. It’s suitable for semi-supervised algorithms, which means it can handle both labeled and unlabelled data.

But there was another issue.

Drug-drug interaction has millions of possible adverse effects, and this could cause doctors and patients to start ignoring the alert. To avoid what the researchers are calling “alert fatigue,” the researchers chose to focus on high priority interactions.

These include interactions that could have life-threatening side effects, disability, or hospitalization.

According to the researchers, analyzing how drugs interact is the first step of the study. The subsequent parts, however, would involve developing and refining the technology that’ll lead to a more precise and personalized result.