Artificial Intelligence and Machine Learning Accelerate Rapidly in FDA Submissions

People working in clinical trials are sometimes criticized as late adopters of technology. However the reality is more complex. The Gartner 2019 CIO Agenda survey clearly outlined the biggest barriers to adoption of AI and machine learning (ML) as staff skills, understanding benefits and uses of AI/ML technologies, as well as data scope and quality.1

In attending a recent workshop hosted by FDA and The Center of Excellence in Regulatory Science and Innovation at the University of Maryland, “Application of Artificial Intelligence & Machine Learning for Precision Medicine”, it was apparent that our industry is getting past those barriers. Across the life sciences and technology sectors, we’ve been bringing in and building the right skill sets, collaborating across specialities, working through challenges like acquiring diverse data to better inform model building, and are generally reaching a tipping point wherein clinical research teams are delivering and utilizing AI/ML technologies to benefit sponsors, investigator sites and patients alike.

As evidence of the recent rise in benefit from the use of AI/ML technologies in clinical research, there has been a skyrocketing number of submissions to FDA that include AI/ML. These submissions grew from just a few to a handful from 2016 to 2020 and then they leapt to over 130 in 2021 and well over 150 in 2022. At the workshop, FDA’s Hao Zhu shared these updates to data published in Landscape Analysis of the Application of Artificial Intelligence and Machine Learning in Regulatory Submissions for Drug Development From 2016 to 2021.2 The AI/ML submissions represented a wide range of technology uses, including solutions like those offered by AiCure to enhance adherence to drug regimen, and deploy digital biomarker assessment and predictive data supporting patient management strategies.

Reflecting on the data Hao Zhu shared, this rapid uptick matches the experiences we’ve been witnessing inside our own customers over the last year. Life sciences teams have been asking well-informed questions, garnering a deeper understanding of AI/ML solutions and driving better results with them. The FDA’s “Good Machine Learning Practice for Medical Device Development: Guiding Principles" has garnered more attention, more quickly than similar technology guidelines in past decades. One of its principles upholds the need for multi-disciplinary expertise, and collaborative engagement across specialties came up frequently during the FDA workshop as a best practice.

Unsurprisingly, we are also seeing more informed discussion of diversity in AI and great lines of questioning around avoiding bias and managing data sets for optimal application during both training and testing. At the FDA workshop, the concept of diversity was articulated by Luca Foschini, President and CEO of Sage Bionetworks, who spoke on “Challenges in AI/ML for Health: Bias, Generalizability, Privacy.” Sage Bionetworks is a nonprofit health research organization that is speeding the translation of science into medicine.

Beyond the workshop, prominent industry voices have surfaced similarly supportive data that barriers have been broken and adoption is in full swing. Gartner’s own 2023 CIO Agenda Insights for the Life Sciences Industry cites AI/ML technologies as having the steepest increase in 2023 funding versus 2022. The Gartner survey shows 51% of life sciences CIOs planning to increase AI/ML funding this year, up from 30% in 2022. And 94% of CIOs see AI/ML as most likely to be implemented by 2025.

It’s already March 2023, where are you on your adoption curve? Contact us now to use already-proven AI/ML technologies to increase patient retention, actively manage adherence in real-time or leverage digital biomarkers to advance your clinical development efforts.

1 Laurence Goasduff, (2019) Gartner 2019 CIO Agenda survey,

2 Liu, Q., Huang, R., Hsieh, J., Zhu, H., Tiwari, M., Liu, G., Jean, D., ElZarrad, M.K., Fakhouri, T., Berman, S., Dunn, B., Diamond, M.C. and Huang, S.-M. (2022), Landscape Analysis of the Application of Artificial Intelligence and Machine Learning in Regulatory Submissions for Drug Development From 2016 to 2021. Clin Pharmacol Ther.