AI’s latest job: Helping NIH fund the nation’s scientific agenda
Illustration: Getty Images
New artificial intelligence (AI) solutions from Leidos are helping the National Institutes of Health (NIH) vet the huge number of grant applications it receives every year.
Why you should know: NIH is the largest public funder of biomedical research in the world. By choosing which research to fund, NIH effectively sets the nation's scientific agenda for years to come.
NIH-supported researchers have won or shared 99 Nobel Prizes in various fields including physics, chemistry and medicine.
The challenge: In 2021 alone, NIH received nearly 90,000 grant applications, making the competitive scientific review process a major responsibility.
The NIH’s Center for Scientific Review (CSR) uses a detailed process to ensure each application receives fair and expert analysis.
It’s such a large and complex job that it requires roughly 20,000 reviewers and 1,300 review meetings each year.
The solution: The center has now adopted two Leidos-developed tools to transform the process:
- The Assisted Referral Tool (ART) uses machine learning to analyze the scientific content in each grant application and recommends a suitable panel of scientists to evaluate the entire application. ART can process 9,400 grant applications monthly, reducing the clerical workload by 80%.
- The Science Similarity Detection (SSD) tool helps reviewers avoid redundancy by identifying applications with similar scientific content.
From the sources:
- “The biggest value of these tools is that they allow CSR review officers to focus on what they do best in order to achieve their mission. These tools result in substantial time savings for the application processing cycle," says Leidos Sr. Research Scientist Dr. Ning Yu.
- “Increasing efficiency is important, but our larger goal is to support funding of the most promising science," says Leidos Sr. Program Manager Ajit Menon. “In this way, taxpayer money will be maximized to achieve the best outcomes for society."
Looking ahead: Menon said his team is developing several new AI solutions to make the process even more efficient, including tools to automate sentiment analysis and quality control.
Please contact the Leidos media relations team for more information.