Office of Technology Transfer – University of Michigan

BlackBox ICU

Technology #6011

Multiparameter patient monitors are defined as monitoring systems that measure and display waveforms and numerical data for patient parameters that are used in the diagnosis and treatment of patients in various healthcare settings. In the modern era of healthcare, use of methods like telemetry and evasive monitoring are becoming exceptionally commonplace. These monitoring systems display multiple input signals from various kinds of vital sign-sensing devices, such as heart rate, respiratory rate, temperature, invasive and non-invasive blood pressure, end-tidal CO2, electrocardiograms (ECG), and arterial oxygen saturation. Thousands of different waveforms are distributed from monitors on a daily basis through a hospital, and clinicians perform waveform analyses on these signals. This data also has a role in research, quality improvement, and malpractice proceedings but typically the vast majority of such information is lost because it is not logged for anything larger than the time the data exists on a local device. The existing technology on collection of medical center waveform data limits the development of automated waveform interpretation efforts and does not provide a complete archive of all patient admission waveforms. As a result, research that explores waveform repository development to permit easy acquisition and storage of medical data can have an important impact in Multiparameter Patient Monitoring Equipment Market that is >1,075M in revenues.

Waveform Depository Development Using Real time Waveform Capturing and Processing for Advanced Patient Monitoring Solutions

Researchers at the University of Michigan developed a waveform capture architecture with the goal of archiving waveforms from patients to incorporate data from all monitored locations in a medical center. This deidentified data can be housed internally and subsequently linked to both an anonymized research data warehouse and an identified medical records system for future research. Additionally, the ability of interface servers to process incoming data in real-time allows the study of physiologic state detection algorithms and alarms in prospective and potentially randomized controlled trials. The resulting waveform depository can be used to develop advanced monitoring solutions from retrospective analyses with a variety of machine learning techniques. In contrast to prior technologies, the system developed at the University of Michigan seeks to capture all potential waveform data on all patients admitted to a single quaternary care medical center.

Applications

  • waveform research in acute and operative settings
  • medicolegal defense mechanism
  • benchmarking software solution in acute and operative settings

Advantages

  • capture of waveforms from multiple nodes on a single network
  • creates tracings linked to a single individual for their entire stay
  • enables real-time processing of waveforms