Artificial Intelligence in the Pharmaceutical Industry May Revolutionize Medicine

February 26, 2021 • Zachary Amos


The introduction of artificial intelligence in the pharmaceutical industry has changed healthcare forever. Artificial intelligence (AI) has the power to scan through massive amounts of data, uncovering subtle correlations and relationships as it goes. 

This analytic ability makes AI an extremely powerful tool for pharmaceutical companies that want to discover new drugs or improve the functionality of their medical devices.

As a result, the pharmaceutical industry is beginning to use AI to develop new treatments. The use of this tech may soon change how the industry approaches medicine altogether.

A Vaccine Designed by AI

In 2019, a team of Australian researchers announced that they had developed the first human drug in history to be designed solely by an AI, a new flu vaccine.

This particular flu vaccine has a newly discovered adjuvant — a compound that boosts the human immune system, helping it to fight off the virus.

The team accomplished this by first training the AI. They used data on existing adjuvants to learn about how they were structured and the kind of properties they have. 

Then, the team generated a massive database of new potential adjuvants. The first AI analyzed these compounds, using its model of the relationship between structure and characteristics to determine which adjuvants were most likely to provide a beneficial, immune-boosting effect.

The researchers then manufactured vaccines with one potential adjuvant and tested their effects. The results were excellent. In some cases, the new adjuvant was even more effective than some currently used in flu vaccines.

The vaccine isn’t available yet — it still needs to go through lengthy human trials. However, the experiment’s success is an excellent sign for the use of AI in drug discovery. If researchers can use the tech to uncover new adjuvants, it may be able to find alternatives for other drugs — and rule out non-viable options sooner.

AI tech could, as a result, speed up drug development. Treatment R&D is a major expense in both time and money. New drugs cost $2.6 billion on average to develop and take an average of ten years to go from discovery to approval.

If AI-designed treatments are viable, they could be a major boon for pharmaceutical companies concerned about the significant investment needed for a new drug.

AI-Powered Medical Devices

Doctors may soon be able to use artificial intelligence to to identify rare or difficult-to-diagnose conditions.

Bayer and Merck & Co., for example, are working on a new AI-powered software tool to help clinicians make treatment decisions about chronic thromboembolic pulmonary hypertension (CTEPH), a disease associated with lung embolism.

CTEPH is often difficult to diagnose. The symptoms of the disease are very similar to other conditions that affect the lungs, like asthma and COPD.

If left untreated, however, CTEPH can quickly become life-threatening. Blood clots from elsewhere in the body can travel to the lungs. There, they may fail to dissolve and instead create scar tissue that clogs up the small blood vessels there.

The team design the tool to help radiologists, who typically diagnose the condition by looking at computed tomography (CT) images of a patient’s lungs. The AI analyzes these images and uses information about the patient’s medical history to identify potential signs of CTEPH.

If effective, the software could make it much easier for radiologists to diagnose the rare condition. This would help patients with the disease get treatment as soon as possible. 

The tech is similar to other AI tools designed to help radiologists analyze CT scans. 

For example, one team of researchers were able to design an AI algorithm that could detect signs of acute neurological illness — such as stroke, hemorrhaging and hydrocephalus — faster than radiologists.

Tech giant Google has also developed similar tech. In 2019, the company assembled a research team whose AI was able to beat radiologists in finding signs of breast cancer.

How Artificial Intelligence May Transform the Pharmaceutical Industry

So far, much of this AI tech is still experimental. Even with the faster development that artificial intelligence can offer, treatment R&D is a slow process. 

Still, it’s likely that we’ll begin to see the impacts of AI on the industry soon — possibly within just a few years. Major pharmaceutical companies and outside organizations are adopting AI at a rapid pace and finding ways to use the tech to transform how treatments are designed and delivered.