New AI developments in healthcare

On LegalAIR, we recently devoted another blog to a new tool for the application of Artificial Intelligence (AI) in healthcare, designed in collaboration with the Ministry of Health, Welfare & Sport (Ministerie van Volkgezondheid, Welzijn & Sport, VWS) and the UMC Utrecht. As it turns out, developments in the field of artificial intelligence, are anything but standing still in healthcare. For example, in early February, the Canadian software company DNAstack launched a new network called Viral AI. This blog briefly explains exactly what this new network entails, and whether a similar development is also upcoming in the Netherlands.

THE VIRAL AI NETWORK

Viral AI is a network that helps track mutations in infectious disease genomes and conduct infectious disease research. Viral AI was created, among other things, to share health data more quickly at the international level. In this way, for example, scientists can have global, representative data sets that are needed in the fight against COVID-19 and any future infectious diseases.

Using this network, genomic, clinical, administrative and other data can be analysed and shared in an innovative way. Previously, only a centralized model existed, uploading data to a single database. Viral AI, on the other hand, is structured in a way that you can connect, analyse and share data without having to move the data in the process. As a result, not only does data management take place in a faster and more efficient way, but the tool is also more compliant with Canadian laws and regulations.

DEVELOPMENTS IN THE NETHERLANDS

This Canadian initiative raises the question of whether a similar development has already begun in the Netherlands. During the COVID-19 pandemic, a host of AI tools were deployed here to map the spread of the virus or the course of the disease. To develop those tools, health data is used, which therefore includes sensitive personal data under the General Data Protection Regulation (GDPR). When developing or applying AI for the purpose of COVID-19, certain rules may require more attention. For this reason, the Dutch Ministry of Health has published a guide called the Wegwijzer AI en Corona, which builds on the general Wegwijzer AI in healthcare. The new guide attempts to provide insight into the specifics involved in developing and applying AI in one place. It is emphatically not intended as technical support for developing AI. Thus, unlike Viral AI, it is not an initiative to achieve more effective Artificial Intelligence applications, but rather a guarantee that the deployment of AI in the fight against the coronavirus is done responsibly.  

At the time of writing, the Dutch Ministry of Health, Welfare and Sport is also working on a law requiring healthcare providers to share data with each other electronically. This is the Bill on Electronic Data Exchange in Healthcare (Wetsvoorstel elektronische gegevensuitwisseling in de zorg, Wegiz). This law aims to make electronical data exchange in healthcare mandatory, so that better quality care can be provided. The law also aims to put an end to physical exchange of data between healthcare providers. What type of health data can be brought under the Wegiz is still to be determined. In time, the Ministry of Health, Welfare and Sport also intends to set requirements for the language and technology used in data exchange, so that they will meet certain standards of information technology products. In doing so, however, the Wegiz does not create a new basis for the exchange of sensitive personal data. Therefore, the GDPR remains the starting point when it comes to the grounds needed to process data.

CONCLUDING REMARKS

Using a renewed infrastructure, the ViralAI network offers a new application of AI in healthcare. AI is also widely used in the Netherlands, both in general health care and in the management of coronavirus. The Wegizmay have the potential to foster such development if health data is increasingly exchanged electronically and further requirements for language and technology will be formulated. One major difference with Viral AI is that the Wegiz is only committed to sharing health data on a national level. Perhaps in the future a system similar to Viral AI could be developed at the EU level, in order to obtain a more representative dataset of infectious diseases. An important condition is that the sharing of health data should only take place if there is a ground for processing in the GDPR. This could therefore potentially create barriers to the dissemination of health data or the development of AI tools at the international level.