24 Nov 2021
Post-doctoral researcher receives prize for best ongoing research
Aaron Ceross’ work on improving navigation of medical device regulations is given an award by the ICEGOV international conference
Post-doctoral research assistant Aaron Ceross works in the Natural Interaction Lab, led by Prof Jeroen Bergmann, and is currently finishing his DPhil in Computer Science. Ceross’ work centres on using machine learning techniques to assist with the complicated world of medical device regulations.
Ceross recently attended the International Conference on Theory and Practice of Electronic Governance (ICEGOV), an international conference on developments in governing electronics. Universities, research centres, governments, industries, and international organizations come together to share their knowledge via publishing and presenting academic research papers.
“We are very pleased to have our work recognised by the wider research community in this space”
As well as spanning institutions, the conference is multidisciplinary, attracting work from computer scientists, engineers, social scientists, and government policy experts. This year was the 14th annual conference and was held in Athens from 6 to 8 October in a hybrid form; having both in-person and virtual attendees.
On winning the award, Ceross says “We are super pleased to be awarded the Best Ongoing Research Paper prize, from a selection of 41 papers in this category. We are very pleased to have our work recognised by the wider research community in this space.”
Ceross explains the motivation for the research, “Medical devices are governed by complex regulations which ensure the safety of the device. The regulations categorise medical devices into distinct risk categories, the effect of which entails differing obligations for safety criteria and post-market surveillance to monitor the risk to patients.”
“Our tool was able to provide highly accurate and precise medical device classifications based on simple descriptions of the device and its intended use”
“Our research looks at finding ways by which to enable innovators to navigate medical device regulations in order to increase compliance and reduce costs. We evaluated the complexity of the regulations as well as online tools, such as one provided by the Australian medical device regulator, the Therapeutic Goods Authority, to make adherence to the regulation easier. “
Current tools for navigating medical device regulations exist, but still prove complicated, as Ceross continues, “We found that the tool provided by the regulator did make navigating the rules around device classification somewhat easier, but was constrained by a lack of explainability of the classification outcome. We proposed a machine learning tool by incorporating information provided to the regulator to sidestep the complexity of the rules. Our tool was able to provide highly accurate and precise medical device classifications based on simple descriptions of the device and its intended use. “
“The next steps are to further explore new ways of integrating principled data science into regulatory analysis such that we can continue to provide new insights and tools for innovators in medical device development. This current research will be integrated into the Natural Interaction Lab’s Oxford Global Guidance tool (OGG). OGG is an ongoing project which provides a suite of tools that facilitates understanding and navigation of the medical device regulation space”
“I moved into engineering because I want to build tools that help innovators”
Ceross has studied a wide range of subjects, which bring him to this particular field of engineering, “My undergraduate degree is in law, but I had always been interested in way by which technology can help bring new insights to legal and regulatory processes. I pursued a masters in computer science and then joined the CDT [Centre for Doctoral Training] in Cyber Security at Oxford. I am now finishing my DPhil in the Department of Computer Science.”
“I moved into engineering because I want to build tools that help innovators better incorporate rules and regulations regarding safety, security, and privacy into their devices. It’s been really exciting to draw on all the different disciplines in order to pursue this goal.”
Ceross concludes, “This is a novel area of interdisciplinary research that we hope will create effective tools, to support innovators from both industry and academia, for bringing new medical devices more quickly and efficiently to market. “