Artificial intelligence (AI) is becoming more and more the subject of daily conversations. This technology, inspired by the functioning of the human brain, fascinates many, worries others while keeping many unaware of its concrete applications. Its possibilities are as numerous as they are diversified, as are the sectors in which it can be used. To help you get a head start on AI, here are three innovations, all from Quebec, that could each revolutionize their respective fields!
Health: An algorithm to target clinical trials in Alzheimer’s disease
Every 56 seconds, a Canadian is diagnosed with Alzheimer’s disease. By 2050, 1 million people in Canada will be affected by this degenerative brain disease, the majority of them women, due in part to their higher life expectancy than men. Alzheimer’s is a disease that progresses slowly and silently, and by the time it is diagnosed, it is already at a very advanced stage. Despite the hundreds of clinical trials conducted over the past twenty years, this dementia remains irreversible, as no cure has yet been found.
If age is considered as its first trigger factor, we always try to identify the molecular factor or factors that would play a key role in its development. Until recently, researchers faced a major obstacle in clinical trials: the inability to predict which individuals in a cohort of participants with family predispositions and mild cognitive impairment would develop dementia and, if so, how long it would take.
However, researchers at McGill University’s Douglas Mental Health University Institute have used Artificial Intelligence, specifically automatic learning (see Artificial Intelligence: From Manual Programming to Deep Learning), to create an algorithm that can predict with unprecedented accuracy to 84% that among people over 60 years of age with memory problems—but who do not prevent them from working—will develop dementia two years later (Rosa-Neto and al., 2017). In an interview with Le Devoir, the study’s lead co-author, Pedro Rosa-Neto, explains how their innovation works: “With only the data provided by imaging their brains [revealing the level of amyloid-beta, a toxic protein that accumulates earlier and more pronounced in people with Alzheimer’s disease], our software [including the algorithm developed in the laboratory] can determine if these people will develop dementia two years later.”
This will have a definite impact on the progress of research in testing the effectiveness of potential treatments. “If we conduct clinical trials specifically with patients who are certain to develop dementia two years later, we will obtain much more probative and conclusive results on the effectiveness of the experienced treatment,” says Dr. Rosa-Neto.
Education: Nao, the teaching assistant robot
A few students from Quebec have had the opportunity recently to receive a regular classroom visit from a very special teaching assistant: Nao, the little robot! Comprising a camera, sensors and microphones, this humanoid robot is also equipped with artificial intelligence, making it capable of interacting with humans, either to answer their questions or to decode some of their emotions. It was Thierry Karsenti, Canada Research Chair on Information and Communication Technologies (ICTs) in Education, who, after discovering Nao in France in 2016, had the idea to test its potential for educational purposes with his team from the Université de Montréal.
The researchers conducted their project with elementary and secondary school students with learning disabilities, as well as a group of elementary school students with autism spectrum disorders. In addition to helping these young people improve their academic skills, particularly in mathematics and reading, Nao’s interventions have helped them develop their communication skills and gain self-confidence (Karsenti and Bugmann, 2017). Autistic students who did not communicate with each other began to do so in the company of the friendly little robot (Découverte, Radio-Canada).
These school-based experiments with Nao were, therefore, “a real success,” according to the team of researchers, and Mr. Karesenti, who advocates that AI becomes a priority in education in Quebec, suggests that it should be perceived differently. “Instead of considering artificial intelligence in education as something magical, it should be seen as a tool with great potential that must be used pedagogically, starting at the elementary level,” he says.
Veterinary medicine: Smart earplugs for healthy horses
As one of Quebec Science’s 10 Quebec inventions of the year 2019, intelligent earplugs for horses are becoming a new tool in the equine preventive health kit. Their inventor, who initially designed these plugs to protect better horses from noise, had the idea of equipping them with a gyroscope, accelerometer, GPS, but also biometric sensors, i.e. sensors that provide data on the horse’s state of health (e. g. changes in temperature or heart rate).
By integrating artificial intelligence into these plugs, this technology allows the biometric data of the horse to be analyzed using equine health data from existing scientific literature—in which AI has been trained in advance. The AI thus provides horse health forecasts that can detect anomalies or even issue an alert to report an urgent health problem if necessary. At the moment, the data is being analyzed in cloud, but the aim is to have it analyzed in real-time, in the plug.