Researchers from Children's Hospital of Philadelphia and Drexel University have created a machine-learning tool that can analyze 53 million patient notes to assess if patients have any similarities in their medical histories.
Researchers used the tool dubbed Arcus to analyze notes from more than 1.5 million patients' data, who had a variety of diagnoses and syndromes, and said it could effectively analyze phenotypes as well as medical similarities among patient populations, according to an April 6 press release from Children's Hospital of Philadelphia.
Arcus was able to identify 9,477 distinct phenotypes.
"The algorithm we developed in this study has the potential to be utilized in finding similarities between clinical trajectories and identifying novel genetic causes of diseases," said Ingo Helbig, MD, pediatric neurologist at Children's Hospital of Philadelphia. "This will allow us to use machine learning in tandem with existing methods to analyze risks and patient prognoses in a more efficient manner at large scale."
The findings were published in the journal Artificial Intelligence in Medicine.