High-Confidence Medical Devices: Cyber-Physical Systems for 21st Century Health Care (Feb. 2009) (full-text).
This report presents the perspectives of the senior scientists of the NITRD Program's High Confidence Software and Systems (HCSS) Coordinating Group (CG), with input from experts from other federal agencies, on the R&D challenges, needs, and strategies for developing and deploying the next generations of high-confidence medical devices, software, and systems. HCSS agencies whose missions are not medical device-specific have found it beneficial to partner in this area because medical device research challenges are similar, if not identical, to those within their purview.
Digital technologies are increasingly being assigned high-level control over the monitoring, sensing, actuation, and communications of medical devices — often with human life in the balance. Through this report and associated HCSS-sponsored national workshops, the HCSS agencies are seeking to illuminate fundamental scientific and technical challenges that must be addressed before we can design and build high-confidence medical devices, software, and systems that operate flawlessly from end-to-end. The report authors sought to paint the landscape of the evolution of medical device technology and the federal investments that have benefitted medical device R&D over time.
Key findings Edit
The authors noted a number of key findings:
- Today's medical device architectures are typically proprietary, not interoperable, and rely on professionals to provide inputs and assess outputs; "families" of such devices also tend to be stove-piped and not interoperable with other "families" of devices.
- In the frequent circumstance that a patient is connected to multiple devices at once, such as in an operating room, clinicians now must monitor all devices independently, synthesize data, and act on their observations, which can be affected by stress, fatigue, or other human factors.
- Medical device architecture is beginning to include wired and wireless interfaces to facilitate networked communication of patient data. But ad hoc efforts to aggregate data across devices designed to operate separately can lead to unintended or accidental results.
- The growing interest in such capabilities as home health care services, delivery of expert medical practice remotely (telemedicine), and online clinical lab analysis underscores the central role of advanced networking and distributed communication of medical information in the health systems of the future. Increased R&D focus on the specialized engineering of networked medical device systems is needed.
- Neither past nor current development methods are adequate for the high-confidence design and manufacture of highly complex, interoperable medical device software and systems ("intelligent" prosthetics, minimally invasive surgical devices, implants, “operating room of the future”), which in years to come will likely include nano/bio devices, bionics, or even pure (programmable) biological systems.
- Today's verification and validation (V&V) efforts are driven by system-life-cycle development activities that rely primarily on methods of post-hoc inspection and testing; these approaches are inadequate in the face of the diversity and complexity of components and interactions in emerging medical devices and systems.
- Today, scientific principles and engineering foundations are lacking that could enable both the design and assurance of high-confidence medical device cyber-physical systems.
Based on their findings, the authors drew the following conclusions:
- Clearly, there is a need for rationally designed high-confidence medical device cyber-physical architectures; a strategic focus on R&D in compositional modeling and design is needed to address the open systems needs, respond to technological innovation, and bridge the jointly cyber and physical aspects of this complex systems problem.
- An open research community of academics and medical device manufacturers is needed to create strategies for development of end-to-end, principled, engineering-based design and development tools. Certifying component devices is necessary, but not sufficient; a key area of research needed is the incremental certified composition of certified components.
- Manufacturers will need access to open, formally composable V&V technology that relies on computational models unifying cyber and physical systems to help establish sufficient evidence. A key V&V research challenge is to understand what is meant by the term “sufficient evidence,” its properties, and how this can be accepted in the global economy.
- The HCSS group recommends that a strategic R&D focus on high-confidence networking and IT for the design, implementation, and certification of open medical technologies be undertaken, both to meet the goals of cost-effective, improved patient care and to spur innovation that promotes U.S. leadership in biomedical technology.
- To enable the necessary holistic cyber-physical systems understanding, barriers must fall among the relevant disciplines: medicine, discrete and continuous mathematics of dynamics and control; real-time computation and communication; medical robotics; learning; computational models and the supporting systems engineering design, analysis, and implementation technologies; and formal and algorithmic methods for stating, checking, and reasoning about system properties.
- Incentives are needed to enable effective cooperation between government, industry, and academia to build the underpinning standards and networking and information technology frameworks (e.g., testbeds) for developing open, interoperable medical cyber-physical systems.