Medical Cyber-Physical Systems
CPS applications have allowed researchers to team up with practitioners to better understand the problems and challenges and provide solutions that can be tested in practical settings. This is the case with medical devices and as a result, health practitioners are required to coordinate the use of these devices and ensure the safety of their interactions. Researchers working in this area are developing techniques to enable the certification of these interactions. This chapter discusses the issue at length.
Insup Lee, Anaheed Ayoub, Sanjian Chen, Baekgyu Kim, Andrew King, Alexander Roederer, and Oleg Sokolsky1
Medical cyber-physical systems (MCPS) are life-critical, context-aware, networked systems of medical devices that are collectively involved in treating a patient. These systems are increasingly used in hospitals to provide high-quality continuous care for patients in complex clinical scenarios. The need to design complex MCPS that are both safe and effective has presented numerous challenges, inclulding achieving high levels of assurance in system software, interoperability, context-aware decision support, autonomy, security and privacy, and certification. This chapter discusses these challenges in developing MCPS, provides case studies that illustrate these challenges and suggests ways to address them, and highlights several open research and development issues. It concludes with a discussion of the implications of MCPS for stakeholders and practitioners.
1.1 Introduction and Motivation
The two most significant transformations in the field of medical devices in recent times are the high degree of reliance on software-defined functionality and the wide availability of network connectivity. The former development means that software plays an ever more significant role in the overall device safety. The latter implies that, instead of stand-alone devices that can be designed, certified, and used independently of each other to treat patients, networked medical devices will work as distributed systems that simultaneously monitor and control multiple aspects of the patient’s physiology. The combination of the embedded software controlling the devices, the new networking capabilities, and the complicated physical dynamics of the human body makes modern medical device systems a distinct class of cyber-physical systems (CPS).
The goal of MCPS is to improve the effectiveness of patient care by providing personalized treatment through sensing and patient model matching while ensuring safety. However, the increased scope and complexity of MCPS relative to traditional medical systems present numerous developmental challenges. These challenges need to be systematically addressed through the development of new design, composition, verification, and validation techniques. The need for these techniques presents new opportunities for researchers in MCPS and, more broadly, embedded technologies and CPS. One of the primary concerns in MCPS development is the assurance of patient safety. The new capabilities of future medical devices and the new techniques for developing MCPS with these devices will, in turn, require new regulatory procedures to approve their use for treating patients. The traditional process-based regulatory regime used by the U.S. Food and Drug Administration (FDA) to approve medical devices is becoming lengthy and prohibitively expensive owing to the increased MCPS complexity, and there is an urgent need to ease this often onerous process without compromising the level of safety it delivers.
In this chapter, we advocate a systematic approach to analysis and design of MCPS for coping with their inherent complexity. Consequently, we suggest that model-based design techniques should play a larger role in MCPS design. Models should cover not only devices and communications between them, but also, of equal importance, patients and caregivers. The use of models will allow developers to assess system properties early in the development process and build confidence in the safety and effectiveness of the system design, well before the system is built. Analysis of system safety and effectiveness performed at the modeling level needs to be complemented by generative implementation techniques that preserve properties of the model during the implementation stage. Results of model analysis, combined with the guarantees of the generation process, can form the basis for evidence-based regulatory approval. The ultimate goal is to use model-based development as the foundation for building safe and effective MCPS.
This chapter describes some of the research directions being taken to address the various challenges involved in building MCPS:
Stand-alone device: A model-based high-assurance software development scheme is described for stand-alone medical devices such as patient-controlled analgesia (PCA) pumps and pacemakers.
Device interconnection: A medical device interoperability framework is presented for describing, instantiating, and validating clinical interaction scenarios.
Adding intelligence: A smart alarm system is presented that takes vital signs data from various interacting devices to inform caregivers of potential patient emergencies and non-operational issues about the devices.
Automated actuation/delivery: A model-based closed-loop care delivery system is presented, which can autonomously deliver care to the patients based on the current state of the patient.
Assurance cases: The use of assurance cases is described for organizing collections of claims, arguments, and evidence to establish the safety of a medical device system.
MCPS are viewed in a bottom-up manner in this chapter. That is, we first describe issues associated with individual devices, and then progressively increase their complexity by adding communication, intelligence, and feedback control. Preliminary discussion of some of these challenges has appeared in [Lee12].