New paper outlines a roadmap for machine learning in clinical care

A group of physicians, including Vincent Liu, MD, MS, research scientist at Kaiser Permanente Division of Research in Northern California, recently published a research paper in Nature Medicine focusing on building a roadmap for the responsible use of machine-learning-based interventions.

To date, adoption of machine learning in clinical care has been limited, the authors write. However, it has the potential to break silos and aggregate a breadth of knowledge across multiple disciplines.

The paper’s authors recommend a framework for deploying machine learning systems in health care that includes the following steps:

• Choosing the right problems
• Developing a useful solution
• Considering the ethical implications
• Rigorous evaluation
• Thoughtful reporting
• Deploying responsibly
• Making it to market

In building a roadmap for machine learning, the paper also urges stakeholders to work together to better understand the nuances and biases that exist in health data before putting forth solutions. Doing so can help promote progress toward reducing social inequalities while improving overall care.

Dr. Liu is also an internal medicine physician with The Permanente Medical Group.

To learn more, read the full paper in Nature Medicine.