Dr. Joris Galland is a specialist in internal medicine and passionate about new technologies. After practicing at the Lariboisière Hospital, he joined the Bourg-en-Bresse Hospital (France). He analyses the trends in medicine as they unfold and which will impact our lives in the decades to come. Originally published by our partners at esanum.fr.
We explained in a previous article that the development of artificial intelligence (AI) depends to a large extent on the collection of a colossal mass of data, known as big data. It is thanks to this data that AIs learn, self-program and improve.
Before the era of Internet and social networks, mass data collection was unimaginable. In the 1990s, a census of the entire French population took several months (and for some components even several years to process) and required colossal human and financial resources. Nowadays, it is done in a matter of seconds.
Nowadays, the Internet and social networks are formidable tools for such mass data collection in record time. In 2018, Facebook's more than two billion active users generated four petabits of data per day, twice the amount of data found in all the books in all US university libraries.1
The creation and collection of this massive amount of data is accelerating year by year, so AIs continue to grow in number and skill. The rise of connected objects and the Internet of Things (IoT) plays a major role in this data revolution.
At the end of the 20th century, access to the Internet was limited to computers. Very quickly the Internet spread to increasingly unusual objects (refrigerators, hoovers, etc.) and increasingly portable objects (tablets, smartphones, watches). These objects are now capable not only of downloading data but also of generating data through sensors and even interacting with the physical world: this is what is known as the "Internet of Things".
The IoT is the natural evolution of technology, an inevitable link between the digital and physical worlds. Since the 2010s, the IoT has taken off with the exponential use of smartphones. By 2020, there was an estimated 50 billion connected objects in the world. Our phones are the biggest generators of data: in France, the average time spent using social networks was 153 minutes per day in 2017, for a daily screen time of over four hours. In 2020, during the COVID-19 lock-downs, this average time reached more than five hours per day. The data world has a bright future ahead of it.
The mobile phone is not the only object that generates data and interacts with our physical environment. The catalogue of connected objects - most of which are single-tasking - is growing all the time. They are omnipresent in our daily lives: watches, living room speakers, cars with driving aids, traffic cameras, sensors for predicting environmental disasters, fire detection, air analysis, home robots for housework, etc. These objects are capable of acquiring data from the environment and from the people who live there. These objects are capable of acquiring signals from the physical world (generating data via sensors) but also of acting on the physical world (triggering an alarm, activating a cleaning robot, etc.).
At present, household appliances and cars are probably the most publicised applications of IoT. But the world of health is not to be outdone, to the extent that patient care and the daily lives of health professionals will be totally transformed within a few years' time.
"Your smartphone will be your future stethoscope"; this was the metaphor I used in 2015 at a conference on e-health.2 In 2020 this prediction will be borne out by the emergence of m-health (mobile health), which includes connected portable objects used by healthcare professionals (pulse oximeters, portable ultrasound scanners, applications, etc.).
The French National Medical Council published a report (2015) to define m-health and specify its different categories: "medical and public health practices based on mobile devices such as mobile phones, patient monitoring systems, personal digital assistants, and other wireless devices".
325,000 health apps are now available on smartphones. In France, these applications were appealing: 43% of surveyed citizens said they used them on a daily basis in 2016, and this figure is most certainly much higher now.3 Furthermore, the world of healthcare has been turned upside down by the arrival of connected objects and surgical robots.
The growing popularity of connected health devices (smartphones, watches, clothing, etc.) has led to the emergence of "quantified-self": self-measurement by the patient in real time. An edifying example is that of connected watches, 46 million of which were sold worldwide in 2018. Half of them were Apple™ Watches, a model that can record heart rate, recognise arrhythmias, perform an electrocardiogram (no less!), assess our physical efforts (calories burned, number of steps, distance, etc.) and even measure our physical activity. Recent models were even equipped with a saturometry function.
There are already examples in the literature showing that the Apple watch has saved lives (and in my opinion is that it will continue to do so). For diabetics, watch prototypes already allow real-time blood sugar readings to be taken in order to adapt insulin therapy as best as possible.
Needless to say, this mass of collected medical information can never be read and interpreted by a single person. It is impossible for a general practitioner to analyse the vital parameters of their 2,000 patients every day, in addition to their consultation activity. An AI will therefore be needed to interpret the data and alert them only in the event of an anomaly.
The collection of physiological data in real time, coupled with the analysis of the data by AI, will undeniably contribute to the emergence of predictive and preventive medicine that will track down the pre-symptoms of a disease. Patients' demands will undoubtedly move in this direction and social security will seize this opportunity because...prevention is cheaper.
The m-health wave is also making its way into the daily lives of health professionals. What doctor has never consulted their smartphone at a patient's bedside or done a medical search on Google? Better still, the smartphone will literally become an ultrasound stethoscope with the arrival of ultra-portable ultrasound probes. The EchoPen4 project promises an ultrasound probe that fits in your pocket, connected via WiFi to the smartphone, with impressive image quality and for only about a hundred euros. Hardly more expensive than a traditional stethoscope.
The arrival of 5G will allow other technological feats. For example, operating on a patient with a robot, remotely and routinely. This feat is already possible, but for the moment it remains exceptional because it requires ultra-secure and expensive connections. Another expected benefit is the accelerated deployment of telemedicine.
The quantified self is therefore a new weapon in the service of health. Patients are becoming increasingly tech-savvy; as consumers of social networks and connected objects, they provide the data needed to perfect artificial intelligence.
This glittering future of medicine comes at a price. The American GAFAMI (Google, Apple, Facebook, Amazon, Microsoft, Intel) and Chinese BATX (Baidu, Alibaba, Tecent, Xiaomi) retain a technological monopoly and collect a large share of the data generated by the IoT. Yet these same companies have a virtual monopoly on AI. Will the medicine of the future be in the hands of IT firms rather than health professionals or specialised companies?