The Chrodis Plus policy dialogue held in Budapest, Hungary on 18 February 2020 has addressed the role of AI and BI solutions in medical decision support solutions.
National screening programs are potentially the most profitable public services in the health system, with the highest per-unit cost. There is a strong social interest in the efficient running of programs, so it is vital to incorporate state-of-the-art IT solutions as widely as possible. During the dialogue, participants looked at two options for information technology:
Artificial intelligence solutions can play a primary role in evaluating the results, supporting medical decision-making and improving the effectiveness of screening programs. Numerous publications are available that corroborate the diagnostic effectiveness of deep learning algorithms, primarily in medical imaging, but also in other medical decision areas. During the dialogue, we assessed which programs could be assisted by AI, what the prerequisites are, how we can create a training database that supports algorithm development, and how we can build an IT infrastructure capable of serving a national program.
The medical application of artificial intelligence also raises new regulatory issues. There are many situations in which artificial intelligence decides whether a person’s condition requires medical attention. For example, remote monitoring of patients with chronic illnesses, where processing large amounts of data without computer support is unthinkable. Regulating the application of AI technology and clarifying liability issues give legislators tasks. AI also poses new challenges in technology assessment procedures. AI as a decision making support for medical staff can contribute to the detection of chronic diseases at an early stage, thus alleviate national healthcare systems from the burden of treating acute chronic diseases. AI is based on extracting knowledge from massive pools of data, which demands the creation of data pools at national or international level. This needs government interaction at three levels:
– IT infrastructure: sound IT systems collecting, storing and distributing data
– Process management: well designed patient pathways that allow an optimal access of patient to screening supported by AI
– Legal regulations: a transparent legal background including data protection regulations, a transparent distribution of access rights to data
Detailed report about the event will be added to this page by 18 March 2020.