AI in Pharma: Key Takeaways From MIT Technology Review

Use Cases & Projects, Scaling AI Benjamin Libman

How is AI being adopted, deployed, and advanced in firms across the life sciences industry? It’s a question whose answer interests everyone from business leaders to IT professionals and data executives — at pharmaceutical companies, healthcare firms, and companies across the greater healthcare and life sciences (HLS) sector. 

In partnership with Eli Lilly and Regeneron, the MIT Technology Review has published a report with Dataiku exploring how AI is gaining momentum across HLS. In this blog, we’ll highlight the three key takeaways from the report and discuss what they might mean for the future of AI in the life sciences.

A Sea Change in Drug Discovery

AI-assisted drug development means patients could benefit from therapies sooner, more targeted therapies could boost success rates, and product pipelines could more easily expand to explore more complex solutions.

Regeneron Pharmaceuticals, for example, already has a dozen different ML and AI models used in its research and drug discovery process—from analyzing large volumes of images to detecting subvisible particles, including proteins and silicon oil droplets, to predicting and preventing potential immune responses to Regeneron’s products. AI augments human researchers, doctors, and other participants in the medical community to make processes more efficient and find information that might otherwise be lost, says Shah Nawaz, chief technology officer and vice president of digital transformation at Regeneron. 

“Wherever we have imaging or a complex workflow, or we have a large amount of unstructured or complex multi- model data, we are seeing significant dividends paid off,” he says. “So, it’s absolutely showing value there.”

Wrangling Health Data Is the Key to the Future

Healthcare AI progress depends on heavily regulated but useful data sets. According to the American Academy of Family Physicians, during the coronavirus pandemic, one in five adults delayed seeking health care or a procedure. Many patients skipped cancer screenings or delayed treatments, leading to an increase in late-stage cancer diagnoses.

AI will radically change treatment of patients and the delivery of health care, especially as more and more patient data becomes available. Between mobile apps that track sleep and blood pressure, provide fitness guidance, and access patient portals, patients leave increasingly larger data footprints that could be used by AI. As much as 30% of data being collected in 2023 is health-related data, and that figure will increase to 36% by 2025, according to global investment bank RBC Capital Markets.

Between virtual health care, remote monitoring, and data- driven AI, patients will increasingly see health care delivered to wherever they happen to be. Patient data augmented by AI can personalize dosages, track side-effects, monitor health metrics, recommend lifestyle changes, and warn of conditions early, says Vipin Gopal, Chief Data and Analytics Officer at Eli Lilly. 

“People would love to get different types of care in their home, and today, that’s largely not possible,” he says. “Next-gen digital technologies, remote monitoring solutions, smartphones and other devices, and AI algorithms powering this ecosystem are largely going to make that happen to a significant extent, where people can get care in the home and have it be very—if not more—effective.”

Consumer interest in tracking health-related information, and remote care becoming more feasible has built momentum in the industry for doctors and medical professionals. All parties will see benefits from the abilities of AI. “We never said, ‘Let’s go and implement AI,’ but instead asked how much of this can be automated, and then what tools and technology do we bring to help augment our process,” says Nawaz. “That approach paid off because we focused on the value, rather than just bringing in the technology for the sake of doing AI.”

Data Is No Longer Just the Purview of Data Scientists

Data science, advanced analytics, and AI tools are quickly extending outside the data science group, into the hands of stakeholders, including researchers, providers, and manufacturers. Highest priority will be easing the burdens of caregivers, and bringing personalized medicine closer to patients.

The health-care industry anticipates a massive opportunity with the rising utility of AI. The World Economic Forum estimates hospitals produced 50 petabytes of data in 2019, and that 97% of that data is not used. ML models and AI systems can use data more effectively, improving the quality and effectiveness of drugs and health care in general. The use of tech promises ways to better understand the needs of patients and doctors.

Accomplishing these goals also means offering the tools to stakeholders—including researchers, providers, and manufacturers—extending the benefits of the most advanced ML models to users outside the data science group, says Kelci Miclaus, global head of AI/ML for health and life-sciences solutions at AI and ML firm Dataiku.

“It’s not just how we accelerate research and new therapeutics, but it’s how we identify the right patients earlier, and how we treat them with the best paradigm and follow that journey in a way that keeps them healthier,” she says. “AI is becoming embedded across the entire value chain, so we’re starting to see this full movement of the connective tissue from molecule to market, for example.”

You May Also Like

Democratizing Access to AI: SLB and Deloitte

Read More

Secure and Scalable Enterprise AI: TitanML & the Dataiku LLM Mesh

Read More

Solving the Ocean Plastic Pollution Problem With Data

Read More

Revolutionizing Renault: AI's Impact on Supply Chain Efficiency

Read More