AI Algorithms Could Predict Rare Disease Risk

 

People within the rare disease community face many barriers in regard to the diagnostic process. It can take years to receive an accurate diagnosis. Paired with medical costs and a general lack of understanding around rare diseases within healthcare, many patients are left without adequate help. According to an article from Medical XPress, researchers from Penn Medicine and University of Florida Health have joined forces to develop artificial intelligence (AI) algorithms that can improve the diagnostic process.

AI Solutions for the Rare Disease Sphere

Called PANDA (Predictive Analytics via Networked Distributed Algorithms for multi-system diseases), this software is being designed to help inform doctors of patients who might be presenting with symptoms of a rare disease. Further, these algorithms could also predict who might develop certain diseases. Ideally, this platform could be used to identify multiple rare diseases. However, at its onset, the predictive tools will specifically be used to identify patients at risk of developing:

  • Psoriatic arthritis (PsA)
  • Granulomatosis with polyangiitis (GPA)
  • Ankylosing spondylitis (AS)

Outside of the above conditions, the initial AI algorithms will also be used for other forms of spondyloarthritis and vasculitis.

Developing PANDA

So how are researchers working to develop this platform? Researchers will source data from PCORnet, which contains de-identified health data from over 27 million patients. Next, the researchers will use machine learning to train the algorithms and enhance predictive power. Finally, the research team will test the algorithms within over ten health systems to increase predictive efficacy and evaluate PANDA. Learn more about machine learning here.

As time passes, and PANDA is used more frequently, researchers hope that this tool could become even more attuned to rare diseases and risk factors. PANDA is not the end-all-be-all. However, it can present a risk factor to patients and clinicians that can be used to improve care. For example, if a patient knows that they have a 30% chance or likelihood of developing GPA, then the patient and doctors can use that knowledge when making care decisions.

Ultimately, artificial intelligence has the potential to greatly change the diagnostic and treatment landscape within rare disease. While PANDA is still in its development phases, it will be interesting to see its impact once fully developed.

 

Jessica Lynn

Jessica Lynn

Jessica Lynn has an educational background in writing and marketing. She firmly believes in the power of writing in amplifying voices, and looks forward to doing so for the rare disease community.

Share this post

Share on facebook
Share on twitter
Share on linkedin
Share on pinterest
Share on print
Share on email