AI-Powered Algorithm Detects Pulmonary Embolisms in EKGs with High Accuracy

Strategic Insights -

Researchers at the Icahn School of Medicine at Mount Sinai have developed an artificial intelligence (AI) algorithm that can detect pulmonary embolisms in electrocardiograms (EKGs) for the first time. This breakthrough could revolutionize the way doctors screen for this life-threatening condition, which occurs when blood clots block lung arteries. The study, published in the European Heart Journal - Digital Health, found that the AI algorithm outperformed current screening tests and may be more effective at determining whether patients have pulmonary embolisms.

Key Takeaways:

  • The AI algorithm, developed by researchers at the Icahn School of Medicine at Mount Sinai, can detect pulmonary embolisms in EKGs with high accuracy.
  • The algorithm fused EKG and electronic health record (EHR) data to improve screening accuracy, outperforming current screening tests and the Wells' Criteria Revised Geneva Score.
  • The researchers estimated that the fusion model was 15-30% more effective at accurately screening acute embolism cases.
  • The model performed best at predicting the most severe cases of pulmonary embolism and maintained consistency across different patient demographics.
  • The study was supported by the National Institutes of Health (TR001433) and involved data from 21,183 Mount Sinai Health System patients.

Statistics:

  • 15-30% increase in accuracy for acute embolism cases detected by the fusion model compared to current screening tests.
  • 21,183 patients from the Mount Sinai Health System were used in the study.
  • The algorithm was tested on EKG data from patients with moderate to highly suspicious signs of pulmonary embolism.

Sources:

  • Somani, S.S. et al., Development of a machine learning model using electrocardiogram signals to improve acute pulmonary embolism screening. European Heart Journal (2021). DOI: 10.1093/ehjdh/ztab101
  • The study was supported by the National Institutes of Health (TR001433).
  • The Mount Sinai Hospital / Mount Sinai School of Medicine.