Office of Technology Transfer – University of Michigan

Detection of severe blood loss using advanced ECG processing algorithm

Technology #6147

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Image Gallery
A raw ECG signal can be decomposed by a spectral transform to reveal frequency information while retaining data on phase and amplitude
Kayvan Najarian
Managed By
Drew Bennett
Associate Director - Software Licensing 734-615-4004
Patent Protection
US Patent Pending

For over two decades, it has been known that metrics like heart rate arterial pressure were not accurate indicators of blood loss. Since these measurements do change substantially until the patient has reached a dangerous level of hypovolemia, a more accurate, real-time method of assessing blood loss has remained a pressing need – particularly in ICUs, battlefields, and other emergency care situations. By decoding signals given off by the heart as it beats, doctors and clinicians can now identify signs of blood loss and hemodynamic decompensation before a patient reaches a critical state and intervene accordingly. This advance will allow for improved treatment and triage as well as streamline healthcare delivery in critical situations.

Decoding the ECG

By applying spectral analysis and vector regression to a raw ECG signal, this algorithm can extract critical time-frequency information while leaving the phase spectrum intact. This rich mosaic of data is then examined for abnormalities in heartbeat that may indicate whether and to what degree blood loss has occurred. Furthermore, an iterative regression can applied to the raw ECG signal to for de-noising. The fitted curve can then be used to determine the location of various landmarks (e.g. the QRS complex) or wireless transmitted at a lower power cost because the data has been compressed.


  • Real-time observation of critically injured patients
  • Assessment of patient recovery from surgery
  • Ambulatory and low-power vital sign monitoring
  • This technique may also be applied to other biological signals as well


  • Low computing and transmission power requirements
  • High fidelity denoising and compression
  • Clinically-actionable information for a previously difficult to diagnose condition