AV¶ÌÊÓƵ

In just a few years, AI is set to revolutionise many areas of medicine, for uses as diverse as image interpretation and data semantics. And primary care medicine is no exception. "AI has the capacity to add very powerful tools to the Swiss Army knife of primary care physicians," says Idris Guessous, vice-dean of the Faculty of Medicine, head of the Division of Primary Care and director of the HUG Innovation Centre, who is overseeing the development of several AI projects in his division. "By relieving them of certain tasks or helping them to analyse all sorts of parameters in an integrated way, physicians will be able to concentrate on their patients while promoting their autonomy. Paradoxically, these tools, which are perceived as disembodied, help to put patients back at the centre of attention.

By improving consultation forms, giving access to verified information, and even helping in management and workflows, AI is working to make primary care medicine more efficient, more focused on patients, and more in tune with the needs and expectations of the population. "These new ways of working will have an impact on training," says Idris Guessous. "AI is being incorporated into curriculum content to prepare the next generation of physicians for the reality of this changing profession, but also into teaching methods. Our students didn't wait for us to come along!"

The healthcare sector is still on the sidelines when it comes to AI tools. However, they represent such advances in medical practice that they are becoming indispensable. These projects are based on existing solutions that are tailored to needs, which means that costs can be kept under control and their implementation accelerated, so that they have a real and rapid impact on both healthcare professionals and patients.

---

AI to provide a better service for walk-in patients

This project, led by Dr. Hervé Spechbach, aims to use AI to improve the anticipation of flows in emergency departments. In outpatient emergency units, the volume of patients fluctuates according to the time of day, the day of the week, the weather, sporting events and many other factors. "Anticipating staffing needs is always a challenge," stresses Hervé Spechbach. "We're always navigating a bit by sight, without always being able to anticipate and adapt resources according to demand.

The project therefore involves using all the data on patient flows, cross-referenced with all the other factors that can have an impact on the number of patients in outpatient emergency departments, so that the AI can predict the flow at 24:00, at one week and then at 15 days in order to plan staffing and other logistical aspects (the need for additional examinations, for example) as effectively as possible. Furthermore, the simple fact of having a view of peaks helps teams to cope with stressful periods, which we know are temporary. The project, if financed, will be tested in outpatient emergency divisions at the HUG.

---

ConfIAnce, the medical information chatbot for chronic diseases

Developed over a two-year period and available to everyone since the beginning of February, "" is the first chatbot for primary care medicine in Switzerland. This innovative and intuitive solution offers reliable answers before, during and after consultations.

It lists around thirty common chronic illnesses such as anaemia, depression or hay fever. For each one, it is possible to ask questions (soon in most languages) and get a reliable answer immediately. "AI thus becomes a complement that can be consulted at any time, leaving more time during medical appointments for essential discussions between doctors and patients", explains Dr. Mayssam Nehme, project leader. "It allows the conversation to continue between consultations by interacting with the chatbot in a fluid way."

---

Optimising teleconsultations

The HUG have already developed a teleconsultation tool, HUG@home. The aim of this project is to add AI tools that will analyse the sound of voices (and perhaps, in time, images of faces) to detect mood changes and signs of stress or depression that could alert both users and the doctors treating them. This project could be aimed at the UNIGE's 30,000 students, who will have access to remote consultations and, if they so wish, feedback on what the tool detects about their psychological state.

"During a face-to-face consultation, physicians automatically analyse their patients' attitudes, movements and manner of speaking," explains Dr. Olivia Braillard, who is in charge of the project. "We hope to be able to do the same in teleconsultation and add a follow-up element for patients. The student population, which is often subject to stress, is looking for such tools, which will help them to manage their mental health more effectively, and for personalised advice to accompany the AI analysis."