Experts discuss: How can AI improve healthcare?

door nexuzhealth op 14.1.2025

AI is a hot topic - and healthcare is no exception. What impact will artificial intelligence have on the sector? What opportunities lie ahead, and what challenges can we expect? We gathered five brilliant minds for an engaging debate: Bart Van den Bosch (IT Director, UZ Leuven), Karlien Hollanders (Patient Expert), Isabelle François (Director Innovation & Strategy, Medvia), Bob Neven (Director Product Strategy, nexuzhealth), and David Clijsters (Data Solution Architect, Cegeka).

NXH_Summit_AI-141

Karlien Hollanders (Patient Expert), Isabelle François (Director Innovation & Strategy, Bart Van den Bosch (IT-Director, UZ Leuven), Medvia), David Clijsters (Data Solution Architect, Cegeka) and Bob Neven (Director Product Strategy, nexuzhealth). 

The challenges in healthcare are significant. On one hand, we face limited resources and staff. On the other, an aging population. Is AI the silver bullet? What is its true potential?

Bob Neven (Director Product Strategy, nexuzhealth): “Artificial intelligence will first and foremost impact efficiency. Many tasks can be performed faster, freeing up more time for patient care. Administrative processes, for instance, have plenty of accessible opportunities. Patient records still require a lot of manual input, but I see opportunities for automation there. Additionally, with all the data available, we can provide better decision support, ultimately reducing readmissions.”
 
Bart Van den Bosch (IT-Director, UZ Leuven): “Personally, I anticipate significant productivity gains in research. There’s an overwhelming amount of medical literature. Tools like Perplexity (an AI-powered research and conversational search engine that answers questions using predictive natural language text) can help us navigate this. However, we must also hope that AI doesn’t generate even more unnecessary articles — nobody wants extra noise. Looking at the central electronic health record (EHR) system developed by nexuzhealth, I see tremendous potential. Imagine correlating an abnormal lab result with a radiology report or a specific pathology — a kind of super-smart index. There’s also untapped potential in image recognition. Medical imaging is improving, but radiologists can’t process everything. AI doesn’t get tired; this colleague always stays focused!”

Isabelle François (Director Innovation & Strategy, Medvia): “From my perspective, I immediately think of project funding for innovative initiatives. Many applications are in diagnostics, image recognition and pattern detection in care pathways. AI might even determine the best care pathway for a patient or support drug development. Another exciting possibility is making medical reports more comprehensible for patients.”
 
The future of digital healthcare lies in better collaboration and data exchange. In her personal story, Karlien shares her vision of the role AI can play in healthcare.
Blog

The future of digital healthcare lies in better collaboration and data exchange. In her personal story, Karlien shares her vision of the role AI can play in healthcare.

Read more

What are the conditions, limitations and challenges of integrating AI into existing healthcare systems?

David Clijsters (Data Solution Architect, Cegeka): “I often compare AI to preparing a delicious meal. You need ingredients (data), a good stove (Graphics Processing Units or GPUs), and a recipe (complex algorithms like large language models). If any of these are lacking, the result won’t be appetizing. This means we need to scrutinize our data. Are there any biases? We also need trust and compliance, as we’re dealing with sensitive information. Risk assessments are essential. Furthermore, transparency around data is critical. Building trust in AI happens gradually, starting with simple use cases.”
 
Karlien Hollanders (Patient Expert): “I think the challenges mainly lie in correlations. What can AI do that we cannot? Finding correlations takes humans a significant amount of time. Patients with rare diseases wait an average of 4.7 years for a diagnosis. If we could quickly identify that patients with certain abnormal blood values also commonly experience X or Y, we could make a difference. That’s where the opportunities — and risks — are. We must be absolutely certain it’s 100% correct.”
 
2409-NXH_Summit_AI-23 (1)

David Clijsters (Data Solution Architect, Cegeka) and Bob Neven (Director Product Strategy, nexuzhealth) 

Bart Van den Bosch: “What AI does today is pattern recognition. ChatGPT or any other AI tool is never aware of the output it produces. When AI begins to reason logically, that’s when we’ll see real breakthroughs.”

Karlien Hollanders: “It’s crucial to keep a ‘human in the loop.’ Correlations still need to be double-checked because they can be completely wrong. AI is so powerful that we can’t verify everything ourselves. It’s a balancing act — how much can the human brain comprehend, and how much freedom are we willing to give AI? That balance must always be maintained.”

Isabelle François: “There are many brilliant minds developing amazing solutions. How do we integrate these into existing healthcare systems? Implementation is essentially a form of change management. Securing the necessary funding is another hurdle. Delivering innovation to hospitals, care homes, home nursing organizations, ... That's the real challenge.”

Bart Van den Bosch: “Hard-coded data is better than text for training AI. Moreover, legal engineering (the field combining legal expertise and technology to improve, automate and digitalize legal processes and services) will be essential. At the same time, we must ensure not everything ends up in the United States, China or India because of overly strict regulations here.”

Bob Neven: “We closely monitor upcoming regulations. Additionally, we stay on top of developments and explore collaborations with specialized partners.”
 
2409-NXH_Summit_AI-21

Dirk Lembrechts, (Manager nexuzhealth School), Karlien Hollanders (Patient Expert) and Isabelle François (Director Innovation & Strategy, Medvia)

Where do you expect the first AI applications in healthcare? What do you personally believe in?

Karlien Hollanders: “I’d love to see AI-generated reports written in patient-friendly language, validated by the treating physician.”

Bart Van den Bosch: “We have an abundance of data at our disposal. Perhaps hospitals could collaborate to leverage this data collectively — for instance, to suggest patterns or assist with image interpretation for triage.”

Bob Neven: “And let’s not forget digital assistants. Just like asking Google about your day each morning, a business assistant in healthcare could provide significant value.”
 
2409-NXH_Summit_AI-01 (1)
In this captivating debate on AI in healthcare, diverse insights emerged: AI can reduce administrative burdens, identify data patterns faster and improve diagnostics. Panelists emphasized the importance of transparency, ethics and human oversight in AI applications. By striking the right balance between technology and human involvement, AI can contribute to a more efficient and patient-focused healthcare system.
 
Topics: Digitalization Events

Bekijk alle

Bekijk alle