Virtual Hospital CPS Test Bed

Medical Internet of Things(MIoT)
Global City Teams Challenge Proposal

Julian M. Goldman, MD 
Massachusetts General Hospital / Partners HealthCare System



We propose to provide access to our Medical Device Interoperability Lab to serve as a Virtual Hospital for the GCTC challenge. The Virtual Hospital test bed would:

  1. Provide remote and on-site access to high-bandwidth streaming data from our simulated hospital environment. Data can be generated by hardware and software simulators and medical devices in our Medical Device Interoperability Lab in Cambridge, MA, and accessed using our open source codebase, which also offers control of devices
  2. Enable data, simulated devices, or data processing algorithms from other CPS test beds to augment our Labs’ capabilities using our standards-based Integrated Clinical Environment (ICE)
  3. Provide access to clinicians serving as remote clinical consultants, or as a “Virtual Doctor”, to answer medical questions related to incoming simulated medical data

 Example Use Cases

  1.  A GCTC MIoT test bed external to our own could remotely read noisy clinical (physiological) signals from a simulated patient in our lab, process the data to improve its quality, and present the enhanced data back to our lab’s Virtual Doctor for interpretation. This scenario simulates analyzing clinical signals to improve medical care in austere environments, including motor vehicles (e.g. automobiles, ambulances) and aircraft.
  2. A research team developing novel clinical information displays could remotely control software and hardware simulators in our Lab and obtain the resultant medical device data-stream in real time, thereby enabling low-cost access to expensive equipment for cost-efficient product development.
  3. Novel sensor signals, for example generated by body area network type devices, could be transmitted into our test bed and integrated with other signals in a time-coordinated manner to test the sensors as part of a larger system to accelerate innovation.
  4. The cybersecurity attributes of medical devices and IT networks can be studied via remote monitoring of the effects of malware on device, network, and system performance in our Lab.

Note: These use cases could be implemented in collaboration with institutions such as Penn, KSU, UIUC, NIST, DocBox, and Intel.

Description of our Virtual Hospital CPS test bed:

  1. Lab Environment: 1800 sq feet in Cambridge, MA – Laboratory of the MD PnP research program, supported in part by NSF, NIH/NIBIB, and DoD (TATRC).
  2. Equipment: Diverse medical devices, including ventilators, vital signs monitors, device-specific networks and gateways, telemetry, intravenous infusion pumps, pulse oximeters, and personal health devices with standardized open source “ICE research interfaces”. Physical and software patient-signal simulators enable the creation of sharable data that does not contain protected health information.
  3. Networking: Remotely accessible infrastructure to support emulation of diverse clinical environments, including operating rooms, intensive care unit, hospital rooms, and home healthcare.
  4. Software: Open source software to compose medical devices and apps into an ICE system that interfaces with external systems such as EHRs, order entry systems, and big data analytics.
  5. Connectivity has been implemented via FHIR, DIRECT and CONNECT (validated with a DoD AHLTA test bed); connectivity to the VA VistA EHR via MDWS (Medical Domain Web Services) interface
Submitted to Smart America Challenge November 12, 2013; updated Dec 11 2015