Building Safe and Usable Medical Device Software: A Conversation with Neeraj Mainkar

Cuon London Conference and Neeraj Mainkar's Background

• The Cuon London conference will take place from April 7 to 9, 2025, featuring world-class leaders sharing insights on software development across multiple domains (10s).
• Neeraj Mainkar is the VP of Software Engineering and Advanced Technology at Proo, a medical device company developing a navigational device for orthopedic surgeons (52s).
• Neeraj Mainkar has a PhD in Computational Condensed Matter Physics and 28 years of experience, with the first 14 years in a small defense contracting firm and the remaining years in medical devices (1m15s).
• Neeraj Mainkar's experience in medical devices includes working on neurological devices, infusion devices, informatics for a microbiology lab, software for a surgical robot, and navigation devices (1m48s).

Ensuring Software Safety and Quality in Medical Devices

• Neeraj Mainkar is passionate about creating software that is not only advanced but also safe, effective, and intuitive to use, with minimal cognitive load on the user (2m32s).
• Ensuring the technical quality of a product is crucial, especially in the healthcare space, where a software bug can be life-threatening (3m9s).
• To ensure software safety, medical device companies must follow strict practices and guidelines, such as the FDA's design controls, which involve documenting every step of the development process (3m32s).
• The development process must be repeatable, and companies must apply various tools to test the device extensively, covering different workflows and scenarios (4m14s).
• To ensure a medical device is safe, a well-defined and repeatable process must be followed, with thorough documentation and utilization of various tools to prevent bugs in the software (4m44s).
• The process involves designing software to be as free of bugs as possible, with a strong emphasis on verification and validation, including user group validation to confirm the device serves its intended purpose (5m6s).

Building a Culture of Safety and Responsibility

• The approach is the opposite of the "move fast and break things" mentality, instead prioritizing safety and responsibility (5m45s).
• To instill a culture of safety and responsibility in a team, it's essential to train new engineers on regulations, documentation, attention to detail, and testing, as well as emphasizing the importance of best practices in code development (6m32s).
• Hiring the right people is crucial, with a focus on finding individuals who are exceptionally skilled and experienced in their craft, particularly in the medical device field (7m6s).
• The hiring process can be lengthy and rigorous to ensure the right candidates are selected, and once on board, they must be trained properly and reminded periodically of the importance of their work (7m33s).
• A key aspect of this culture is the "Mom test," which involves considering whether one would be comfortable with a loved one using the device being developed, and if not, taking steps to address any concerns and ensure the device's safety (7m52s).

Hiring and Team Building for Medical Device Software Development

• Engineers working on medical device software come from varied backgrounds, and it's essential to have a diverse demographic, including experienced individuals who have worked on medical devices before and can serve as mentors (8m46s).
• Key positions, such as principal architects, test managers, and development managers, require prior medical device experience, while engineers and testers are hired based on their technical knowledge and attitude towards software development (9m30s).
• Technical knowledge is assessed through tests and small projects to ensure candidates can solve problems and demonstrate their skills (9m53s).
• Attitude towards software development is also crucial, with a focus on a methodical and engineering approach, rather than a rushed or lazy attitude (10m15s).

Software Development Approach and Collaboration

• Medical device software development requires a rigorous and disciplined approach, following a process and ensuring well-defined architecture before coding (11m28s).
• The ability to work well with others, including cross-functional teams, is essential in medical device software development, as it involves collaboration with various stakeholders, including quality, regulatory, hardware, and systems teams (11m58s).
• Understanding the role of software in the larger system and the importance of systems engineering is critical for medical device software developers (12m31s).

Usability in Medical Devices and Its Importance

• In the realm of medical devices, "usable" refers to how intuitive a device is, with the goal of making it as natural and easy to use as possible, requiring minimal training and cognitive load (13m11s).
• A more technical definition of usability involves measuring cognitive load, which can be done through formative testing, where users are given the device with minimal guidance and their ability to quickly pick up on what needs to be done is assessed (13m52s).
• Usability testing involves observing users in a separate room, allowing them to use the device with minimal help, and measuring how quickly they become comfortable using it and how many errors they make (14m8s).
• Designing for low cognitive load and high usability involves making the design user-centric, which requires involving the user from the beginning of the design process, gathering feedback, and making continuous improvements (15m16s).
• Involving users, such as surgeons, in the design process from day one is crucial, and this can be done through mockups, feedback sessions, and continuous feedback groups (15m40s).

Balancing Usability with Other Design Aspects

• Companies, such as Intuitive Surgical, the maker of the da Vinci robot, prioritize usability and intuitive design in their products (13m19s).
• One of the challenges of designing for usability is balancing it with other important aspects of software design, such as performance, to ensure that the device is both usable and functional (16m14s).
• The standard for usability in medical devices is IEC 62366, which provides guidelines for designing and testing usable medical devices (13m48s).
• When designing medical device software, it's essential to strike a balance between usability, maintainability, and performance, as prioritizing one aspect over the others can lead to issues, such as user complaints about slow performance or difficulties in maintaining the user interface (16m54s).
• Keeping surgeons and users in the loop and having them test the device, provide continuous feedback, and give input during the development process is crucial in ensuring the final product meets their needs and is safe and usable (17m41s).

Software Updates, Enhancements, and Postmarket Surveillance

• In the medical device space, software updates and enhancements are handled differently than in other industries, with regulating agencies like the FDA recognizing the need for postmarket surveillance, which involves monitoring the device's performance, addressing issues, and gathering user feedback for continuous improvement (18m37s).
• The FDA requires medical device makers to conduct postmarket surveillance, which includes monitoring device usage, maintaining a catalog of issues, and gathering user feedback to identify areas for improvement and address pain points (19m2s).
• Medical device companies can continually improve their software design and release updates, although not as frequently as non-medical and non-safety-critical software, with a typical frequency of every three to six months (20m14s).
• A feedback loop is essential for gathering user input and addressing issues after the device's release, with channels in place for users to report problems and suggest improvements (20m29s).

The Role of AI in Medical Device Software Development

• AI is being used in the medical device space to create actual products, with companies like the one mentioned in the text developing AI-based products (20m52s).
• AI is making a significant difference in the medical device space, particularly in usability and safety, with developers using AI tools like chatbots to create code and automate unit testing, which can be done faster and more accurately than by humans (21m6s).
• The use of AI in automated unit testing helps reduce human error in software development, which is a major contributor to bugs in software (21m56s).
• There are plans to use AI for virtual reality-based simulation testing, which can help identify usability issues and provide design improvements (22m28s).
• AI tools can analyze user interactions, identify usability issues, and suggest design improvements, and some tools can even predict potential user challenges and create alternative user workflows (22m38s).
• The use of AI in software engineering is still in its early stages, and its potential applications are vast, with the ability to expand the universe of system-level usability testing (23m24s).

Challenges in Medical Device Software Design

• One of the challenges in designing medical devices is catering to a diverse user base, including different user profiles such as surgeons, mettechs, and nurses, each with their own unique needs and requirements (24m5s).
• Striking a balance between catering to different user groups and creating a seamless user experience is an ongoing effort in medical device design (24m44s).
• Another challenge is making the system as usable as possible, which requires careful consideration of various factors, including complexity and user diversity (24m55s).
• Integrating medical device software with older legacy systems can sometimes limit flexibility, especially in complex workflows involving multiple devices and software systems (24m59s).
• The rapid evolution of technology and new tools can be a challenge, as engineers must balance the desire to utilize new technologies with focusing on user needs rather than just incorporating exciting new technologies (25m37s).
• Engineers must control the urge to use new technologies for their own sake and ensure that user needs are the primary consideration, rather than the other way around (26m10s).
• AI-based tools can also present challenges, such as data overload, as modern software often provides a large amount of information to users, which can be overwhelming and confusing if not properly arranged and presented (26m42s).
• Information architecture, data visualization, and contextualizing data in software interfaces have become significant responsibilities for engineers developing medical device software to prevent data overload and ensure useful data presentation (27m24s).
• Engineers must strike a balance between providing users with the right type of data for decision-making and hiding unnecessary data, as suppressing data or providing too much data can be counterproductive (28m11s).

Contact Information and Conclusion

• Neeraj Mainkar can be reached for further conversation at "N Mainkar", which is his first initial and last name (28m44s).
• The audience is invited to reach out with any questions they may have, and a response will be provided (28m54s).