AI System Reveals Hidden Diagnostic Patterns in Electronic Health Records

While EHRs contain vast data, current diagnostic tools struggle to identify subtle links between medical events across time. Now, a new artificial intelligence (AI) system can uncover hidden diagnostic patterns within these records, turning fragmented data into actionable diagnostic insights.

Researchers at the Icahn School of Medicine at Mount Sinai (New York City, NY, USA) have developed the AI system InfEHR (Inference on Electronic Health Records) to link unconnected medical events and reveal previously unseen relationships. Unlike most AI systems that apply uniform diagnostic models, InfEHR tailors its analysis to each patient by building individualized networks from medical visits, lab results, and treatments. This patient-specific “diagnostic web” allows the system to ask adaptive questions and deliver highly personalized clinical insights.

In a study, published in Nature Communications, the system used deidentified EHR data to successfully detect neonatal sepsis without positive blood cultures up to 16 times more accurately than existing methods and predicted postoperative kidney injury 4−7 times more effectively. Remarkably, it achieved this without needing large training datasets, demonstrating its ability to learn directly from patient records and adapt across hospitals and populations.

In addition to diagnosis, InfEHR introduces a crucial built-in safety feature by signaling uncertainty when data are insufficient, ensuring reliability in clinical settings. Its developers plan to expand its use for personalized treatment guidance by integrating data from clinical trials, helping clinicians determine which research findings best apply to individual patients. This probabilistic approach could bridge gaps between controlled research environments and real-world patient care, improving precision medicine delivery.

“Traditional AI asks, ‘Does this patient resemble others with the disease?’ InfEHR takes a different approach: ‘Could this patient’s unique medical trajectory result from an underlying disease process?’ It’s the difference between simply matching patterns and uncovering causation,” said lead author Justin Kauffman, MS. “Clinical trials often focus on specific populations, while doctors care for every patient. Our probabilistic approach helps bridge that gap, making it easier for clinicians to see which research findings truly apply to the patient in front of them.”

Related Links:
Icahn School of Medicine