With healthcare systems around the world suffering from increased workloads and personnel shortages, AI may seem like a wonder drug. Artificial intelligence is here to stay, experts said during Day 2 of Healthcare Week Luxembourg, but it’s crucial to keep patient care and the needs of medical professionals in focus. Discover five key takeaways and the winners of the Healthcare Week Luxembourg awards.
AI is here and it’s already being used – but don’t rely on it too much
Like other medical specialties, radiologists suffer from a number of issues: increased demand for services, staff shortages, increased complexity, high patient volume, and interpreting errors. AI, while not a solution in and of itself, can help, said Emanuele Neri, professor and chair of the radiology department at the University of Pisa. It can reduce artifacts (meaning it minimises distortions and errors in imagery), speed up the acquisition of images, find tumours or fractures, highlight urgent findings, and support decision-making.
But AI brings with it the risk of errors and the loss of skills, he cautioned. Automation bias – the tendency that people have to favour machine-generated decisions in spite of human decisions or data to the contrary, can lead to mistakes.
What’s more, he warned that an over-reliance on AI can lead to de-skilling. It’s like driving with your GPS, said Neri. When was the last time you used a paper map? Do you remember how to read a paper map, or have you become overly reliant on your phone? It’s key for medical professionals, particularly among the younger generation, to retain their skills. The final responsibility lies with the medical doctor, Neri emphasised, a sentiment echoed by other physicians during a panel discussion.
Hybrid healthcare: at the service of vulnerable populations
“A disease never walks into a doctor’s office,” said Markus Oei, a Dutch ENT surgeon with over 25 years of clinical experience. Rather, it’s humans who seek a doctor’s care. It’s important to remember this as AI and digital advances roll into the world of healthcare. Groups with the most medical needs often have the lowest level of digital skills – and healthcare professionals need to keep these vulnerable groups in mind.
“AI will change healthcare, but not solve it,” he argued. The future of healthcare will be hybrid. AI, for instance, can translate texts or generate videos for low-literate patients and non-native speakers, helping them to follow care plans after undergoing surgery.
Developing AI tools: boost collaboration between tech and medical teams
AI tools, whether used for cancer screening, summarising clinical notes, or synthesising data, need to be accessible and easy for clinicians to use. Cooperation between IT experts and medical doctors is key.
“The most important barrier to reinventing healthcare itself is the doctor,” said Jérôme Lechien, professor of otorhinolaryngology and head and neck surgery at the University of Mons, Belgium. There’s a big gap between the engineer designing the AI tool and the doctor who will actually use it, he argued, and doctors are creatures of habit. If it’s too complex, they won’t use the tool. Healthcare workers, therefore, should be at the centre of these processes.
But don’t wait until the end of the design process to integrate feedback from the end user, said Jacques Federspiel, chief information security officer at the Hôpitaux Robert Schuman. Alexandre Hebert, director general of medical affairs at Vivalia in Belgium, noted that medical chiefs are involved from the start of a project. Training, AI literacy, and being on the lookout for “hallucinations” are crucial. “We can have the best tool in the world,” he said, “but if the teams can’t use it in a good way, it’s not a good idea.”
Using AI to increase empathy
Several of the medical professionals who spoke during panels highlighted the importance of focusing on patients and improving their care. That is, after all, why they went into the field of medicine. But due to increasing workloads, reporting tasks, and bureaucratic burdens, there’s less and less time to really interact with patients.
That’s where AI comes in. By efficiently handling routine tasks like taking a patient’s medical history, recording their symptoms, or analysing their voice, especially important when it comes to disorders that affect the larynx, that leaves more time for in-depth and meaningful discussion with a patient, said Lechien. “Most of my colleagues and patients think that AI equals dehumanisation,” said the ENG specialist. “But it’s the opposite. AI will humanise a consultation because the doctor will gain time.”
Data, data, data
The expression goes: garbage in, garbage out. For all of these AI tools to work, clean, standardised, and high-quality data is crucial. “Data is essential and the most important thing in order to have successful AI tools in healthcare,” said Patricia Garcia Canadilla, senior data scientist at Hospital Sant Joan de Déu in Barcelona.
Professionals, patients, and researchers should also be able to share data across borders in a secure and easy manner – something that the European Health Data Space (EHDS) regulation should facilitate in the coming years.
HWL winners: eSklepios, MPC Therapeutics, IEE, and CHU Liège
Finally, the winners of the Healthcare Week Luxembourg awards were announced during the conference. They are: eSklepios (the Luxembourg startup offers a modular electronic record system); MPC Therapeutics (the Geneva-based startup, which is also active in Luxembourg, has created a way to “rejuvenate” immune cells to help them fight cancer); CHU Liège (the public university hospital has developed an advanced genomic screening programme called Baby Detect that aims to find rare diseases in infants, enabling immediate intervention and treatment); and International Electronics & Engineering (IEE, together with Vienna-based Saphenus, has designed smart sensors that are integrated into the shoe insoles of prostheses, improving amputees’ stability and mobility while reducing phantom pains).
Enjoy the pictures below:
