Forum Medical Information Technology 2024

Medical software, embedded or stand-alone, is currently experiencing a boom. Established manufacturers, whose focus in the past has tended to be on hardware solutions, as well as a number of start-ups, are hoping that software solutions will enable them to enter the market more quickly with significantly leaner structures and lower capital expenditure. Further advantages are also seen in the later operating phase, in which it is believed that changes and functional enhancements can be implemented quickly and unbureaucratically thanks to an agile development process.

However, with the publication of the Medical Devices Regulation (EU) 2017/745, this is offset by the risk classification with a risk class IIa or higher, which requires the involvement of a notified body. This body is not only responsible for assessing the conformity of the quality management system and the technical documentation of the product before it is placed on the market, but also ensures that conformity is maintained through regular monitoring. This also includes evaluating all changes to the software before implementation to ensure that the essential safety and performance characteristics are still fulfilled and - in the event of significant changes - initiating a reassessment. Depending on the capacity and workload of the position, such a reassessment can take several weeks to several months.

This is at odds with agile working, which often generates a new revision on a weekly basis. The presentation explains which strategies can be used to bring both worlds together in order to minimize friction.

The European Health Data Space (EHDS) is a draft regulation proposed by the European Commission that is currently being negotiated at EU level. It is expected that co-legislators (Council and European Parliament) reach a political agreement before the European elections in June. The text is quite novel, proposing common rules, standards, infrastructures and a governance framework for the use of electronic health data under what is called the ‘primary use’ purposes (meaning diagnosis and treatment) and the ‘secondary use’ purposes (meaning policy-making, research and innovation). In this talk we intend to present the new legal regime, looking at the challenges and opportunities of the EHDS. We intend to discuss the impact on patient-doctor relationship, referring to the new citizens’ rights in primary use, the priority categories of ‘personal electronic health data’ which should be accessed and exchanged in a specific format, and the obligations for doctors (as ‘data holders’) in secondary use. Digitising health records and creating systems that enable them to be securely accessed by citizens and shared between different actors in the health system and cross-borders has been a long-term objective with many barriers along the way, such as technical, legal, financial, and even societal (element of trust in the system). This regulation will give a new push to health data sharing in the EU, allowing the reuse of health data and fuel the data-driven market.

There is a need for cross-manufacturer data exchange and communication in the healthcare application landscape. This paper proposes a standards-driven strategy from a medical device manufacturer's perspective that takes advantage of multiple healthcare standards to improve product development and quality assurance. By integrating DICOM, FHIR and IEEE 11073 SDC, a cross-manufacturer communication protocol for medical devices, into connectivity strategies, manufacturers can ensure interoperability between devices, applications and systems. An optimized connectivity strategy offers improvements in interoperability, data management and reduced operating costs for medical device manufacturers.

In this talk, we perform a review of the state-of-the-art in machine learning with focus on hybrid and trustworthy systems in medical imaging. We start with a short summary of the general developments of the past in machine learning and how general and specialized approaches have been in competition in the past decades. A particular focus will be the theoretical and experimental evidence pro and contra hybrid modelling. Next, we inspect several new developments regarding hybrid machine learning with a particular focus on so-called known operator learning and how hybrid approaches gain more and more momentum across essentially all applications in medical imaging and medical image analysis. As we will point out by numerous examples, hybrid models are taking over in image reconstruction and analysis. Even domains such as physical simulation and scanner and acquisition design are being addressed using machine learning grey box modelling approaches. Towards the end, we will investigate a few future directions and point out relevant areas in which hybrid modelling, meta learning, and other domains will likely be able to drive the state-of-the-art ahead.