Responsibility for Bespoke 3D Printed Surgical Robots

Clarifying responsibility for 3D printed surgical tools created using AI for individual patients

Project Duration: April 2019 – April 2022

Medical Team Performing Surgical Operation in Modern Operating Room

Credit: iStock/gorodenkoff

Determining responsibility for AI-designed surgical tools

The Challenge

Robot-assisted surgery is one application of robotics in healthcare. Advances in robotics and 3D printing have created the possibility of creating much smaller surgical robots designed specifically as tools for a surgeon to treat an individual patient. Artificial intelligence (AI) offers the means of automating the design process to create a surgical robot that is optimised for the individual patient and the surgical intervention required. This design is then manufactured using 3D printing to create a surgical tool used by a surgeon in the operating room.

This presents both opportunities but also challenges. For example, medical devices usually have a single or limited number of configurations that can be physically tested prior to their approval and adoption. In this case, the tool is unique, and has only been tested in a computer simulation by an AI. The lack of direct human involvement in designing this surgical tool also creates uncertainty over who is ethically responsible if the tool is faulty during an operation, as the design process was automated. Identifying the potential risks and ethical responsibilities at each stage contributes to the development of socially responsible surgical tools aimed at enhancing patient care and recovery.

Responding to the Challenge

CSIRO’s Responsible Innovation and Active Integrated Matter Future Science Platforms are collaborating to advance applied research on the ethics and governance of AI, robotics and autonomous systems. Autonomous Design is advancing new solutions with evolutionary robotics but how we navigate and manage the development and deployment of these opaque complex adaptive systems, along with the proliferation of algorithmic decision-making, autonomous systems and machine learning is a critical pathway to responsible innovation.

Dr David Douglas was appointed as a Postdoctoral Research Fellow in April 2019 to examine and map these systems. His case study on the automated design of surgical robotics aims to establish and test an ethical justification for how the responsibility for the design and implementation of surgical tools might usefully be assigned among various stakeholders. His research involves researchers, software engineers, surgeons, patients, fabricators, and regulators to co-create a map of the process for designing and using these tools that describes the risks and benefits involved, the stakeholders in each stage of the process, what they are responsible for and how.

Project Impacts

This project will demonstrate the effectiveness of adopting a responsible innovation approach in medical robotics by showing how the attribution of ethical responsibility in autonomous design systems can be addressed during development. Incorporating the perspectives of stakeholders allows the practical concerns of those who can affect and are most affected by this new technology to inform its development at an early stage.

Team

CSIRO: David Douglas, Justine Lacey and David Howard

More information

Additional information on this collaboration is available as follows:

CSIRO Active Integrated Matter Future Science Platform: https://research.csiro.au/aim/

CSIRO Active Integrated Matter Future Science Platform’s Autonomous Design Test Bed: https://research.csiro.au/aim/home/aims-research-test-beds/autonomous-design/

Who bears responsibility when AI systems go wrong?

References

Douglas, D.M., Howard, D. & Lacey, J. Moral responsibility for computationally designed products. AI Ethics 1, 273–281 (2021). https://doi.org/10.1007/s43681-020-00034-z