- April 21, 2022
- Research
“Art at the crossroads of several sciences”, medicine at the beginning of the 21st century is about to experience a new revolution, like the one experienced by the world of painting in the 19th century. Artificial intelligence (AI) is changing medical paradigms, just as photography yesterday renewed the foundations of painting.
Medicine is an art as much as a science, it progresses as much thanks to human intuitions as to technical and technological advances. The second is often used to confirm, refine or invalidate the first. Nowadays, medicine is making giant strides thanks to artificial intelligence and its algorithms which make it possible to improve care pathways.
More and more specialists conceive of this as a form of a new medical revolution, in the same way as the appearance of vaccines or penicillin, and a paradigm shift similar to that caused by photography in painting where this new technique had put an end to the realist current and favored the impressionist or surrealist movements. In medicine, the revolution is underway in a wide range of applications.
In terms of diagnostic assistance, artificial intelligence makes it possible to process a set of data and give crucial indications to caregivers. The algorithm can prove to be a valuable auxiliary, in particular thanks to the deep learning method. Thanks to the latter, algorithms are, for example, able to recognize tumors and make recommendations by analyzing a scanner, much more precisely than a human eye, even informed, but above all faster. “Me, a radiologist, I measure the size of the cancer, I describe where the possible metastases are, but it’s a very visual analysis, from which I can’t extract a lot of parameters. While the AI algorithm, by taking the image at the pixel scale and analyzing it without a priori, gives a result much more relevant than what our eye is capable of”, explains Professor Olivier Humbert, the practitioner at the Antoine-Lacassagne Center and researcher at the 3IA Institute, for Nice-Matin. When we know that cancer detected at the earliest is cancer better treated, artificial intelligence could make it possible to drastically improve the life expectancy of patients and limit the loss of chance.
In 2018, an AI, called BioMind, had thus beat down a team of 15 senior doctors from major Chinese hospitals. According to Wang Yongjun, executive vice president of Tiantan Hospital where the experiment was conducted, the training received by the AI enabled him to become competent in the diagnosis of neurological diseases such as meningiomas and gliomas, with an accuracy rate of more than 90%, that of a recognized expert. On less complex but equally strategic issues in a context still marked by the health crisis linked to Covid-19, AI is already fully operational. The American company Remark Holdings, one of the leaders in the sector in the United States, has thus implemented automatic temperature measurement sensors via thermal biosecurity kits powered by artificial intelligence. Already operational in several establishments in New York City, Remark Holdings solutions are also used in Las Vegas to make temperature-taking less invasive.
In terms of medical innovation, the “sky’s the limit” for artificial intelligence
On December 16, the AI proved itself in the detection of dementia, a common neurodegenerative disease, but the onset of which is still difficult to predict for humans. According to a new large-scale study, it could predict which people are at risk of developing dementia two years in advance. Developed by English researchers at the University of Exeter, this AI model was designed based on the analysis of patients, between 2005 and 2015. If no participant had dementia at the start of the experiment, one in ten (1,568) received a new diagnosis within two years of their visit. Warned, the latter were able to be taken care of early, but also to make their arrangements.
Today, specialists in the medical field are very enthusiastic about the contributions of AI to their expertise.
Dr. Ghanimi Rajae valued as well as in the future, “medicine will be more preventive than curative”. Future algorithmic feats should accentuate the personalization of medicine and thereby produce “models (…) capable of helping us (doctors, Editor’s note) to determine the most effective interventions, whether it is medication, lifestyle choices or even simple changes in diet” in order to improve the health of patients. Complete paradigm shift, the medicine of the future will be improved medicine” which is not intended for patients, but for healthy individuals likely to develop a given disease”: no longer need to wait for the symptom to cure the patient.
Another challenge to be met is the training of healthcare personnel, who are still unaccustomed to working with artificial intelligence. Some university courses have already taken the lead. “The algorithms, the doctors must have the capacity to understand them and to criticize them. However, they do not do mathematics or algorithms in their studies. We have therefore set up a university degree to train them in understanding the mechanisms of AI. This course has met with success, with twenty-five enrolled in the first year”, explains Professor Olivier Humbert.
