Prevention of infections in the hospital
Nosocomial infections (acquired in hospital) affect 10 to 15% of the world’s population. What if artificial intelligence could help prevent and control these infections? This is the subject of research by Douglas Teodoro, assistant professor in the Department of Radiology and Medical Informatics at the University of Geneva (UNIGE): “We seek to predict the risk of infection for each new patient , by building models representing his career in the hospital. To do this, we collect data relating to the beds and premises occupied, the staff who have been in contact with the person, etc. The algorithm manages all this history. The pooling of this mass of data generates new information on the characteristics of infected patients. Ultimately, the idea is to be able to help teams introduce new barriers to reduce the risk of contamination.”
Let’s take an example: “You should know that approximately 85% of clinical trials aimed at developing new treatments end in failure before being approved by the European or American drug agencies”, indicates Douglas Teodoro, assistant professor at the Department of Radiology and Medical Informatics at the University of Geneva (UNIGE). In question, among other things, the presence of flaws in the research protocols, written by the various researchers involved and which include all the information on the method, the objectives, the progress, etc., of the study. “We have registers in which no less than 500,000 clinical trials, whether successful or not, are listed. Using algorithms, we can quickly sift through and analyze these databases to assess the risk of failure and identify which parts of a new protocol are problematic.” Improving these processes could bring new drugs to market faster and thereby reduce their price, which depends in part on the costs of research – around 1.3 billion francs for each drug approved.
Regarding the effectiveness of drugs, we know that it can vary depending on the individual. The creation of homogeneous cohorts, using AI, could be useful for predicting the response to treatments of people sharing the same characteristics for the purpose of personalized medicine, explains Douglas Teodoro: “We use algorithms to give a better digital representation of each person, their health characteristics, disease, genetics, background, etc. The variables are so numerous that even the best doctor cannot assimilate them alone.”
Screen for acute renal failure
In the Nephrology and Hypertension Department of the Geneva University Hospitals (HUG), a study on the role of artificial intelligence (AI) in identifying patients at risk of acute renal failure could soon see the day. At the HUG, some 20% of people are affected. Left untreated, this disease has serious consequences: chronic kidney failure, dialysis, lengthening of hospital stay and increased mortality. “Today, to discover it, you have to expressly seek it. We use a scoring system for this. But thanks to AI and its powerful algorithms, we might be able to better identify people at risk throughout the hospital and quickly apply preventive measures to avoid complications, ”explains Pre Sophie De Seigneux, doctor- Head of the HUG Nephrology and Hypertension Department. Several works in other hospitals have shown the interest of AI in these situations.
We wish to give thanks to the writer of this short article for this amazing material
Optimizing clinical research with AI
Explore our social media profiles and also other related pageshttps://yaroos.com/related-pages/