2/05/2020
The United States Government Accountability Office (GAO) released a report, “Artificial Intelligence in Health Care: Benefits and Challenges of Machine Learning in Drug Development.” It outlines six options for policymakers in response to challenges in the use of machine learning in drug development. But why would machine learning be useful in clinical research?
What is Machine Learning?
Machine learning is a computer operation that uses artificial intelligence (AI) to learn and adapt. It applies AI to new data using previous experience to determine the best option. Google uses machine learning to suggest sites you might be interested in reading based on what you commonly search. Another example is when a social media site suggests connections or pages that you may be interested in based upon the history of sites you visit or your listed workplace. Applications in clinical research include assisting in early development of drug development in discovery, preclinical research, and clinical trial design.
How does Machine Learning and AI Help with Clinical Trial Design?
Machine learning can be applied to selection of subjects, randomization groups, and recruitment. Difficulty in subject recruitment is a frequent problem that extends the time needed to accrue data crucial to the outcome. Using machine learning could lead to the development of lower cost processes that increase recruitment and shorten study length. Because it would free resources, these resources could be shifted to other needed areas such as focusing on rare diseases.
Where Are We Now with Machine Learning?
Machine learning in clinical research is an emerging field, so it needs additional research. There is a shortage of high-quality data that is fit for the purpose of AI, in part because gathering appropriate data is difficult due to data privacy concerns.
Machine learning can reduce uncertainty, increase insight to drug discovery, and aid in selection of candidates for clinical research. In turn, it can assist in study design to increase recruitment efforts. This can lead to fewer failed clinical trials, therefore reducing costs and time required for a successful outcome.
You may also enjoy our blog, “Critical Thinking and Artificial Intelligence.”
- The Clinical Pathways Team
Enjoy this blog? Please like, comment, and share with your contacts.