Feb 07, 2024
Is AI the Answer the Healthcare System Has Been Longing For?
Digital transformation, Artificial intelligence
The recent health crisis has confirmed something that experts have denounced for many years: Quebec's healthcare system is deficient, and access to care remains complex. Emergency rooms are overcrowded, hospitals are operating beyond capacity, it’s challenging to find a family doctor, and the waiting lists for surgery are interminable. The network may be "free", but it cannot fail the population like this.
What if I told you there are already existing solutions to these problems? What if I said that technology is now mature and ready to help our system that is struggling under immense pressure? With the help of AI, optimization decisions can be made more efficiently, and robots can easily ease the workload of our overworked practitioners. Wouldn't this be a great solution?
The Commission on Ethics in Science and Technology (CEST) has recently examined the issue and published some confronting results. The digital transformation of the Réseau de la Santé et des Services Sociaux is not advanced enough to integrate artificial intelligence devices into its infrastructure.
I'll give you a second to digest this report.
One of our most prominent institutions, the one that eats up around 40% of Quebecers' budgets, is not sufficiently digitized to integrate the technological tools that could save it. The year is 2023. Where have we gone wrong? Where can we start to catch up?
To implement artificial intelligence systems, you first need to have access to digitized data and ensure that the various interfaces can communicate with each other. This requires abandoning the traditional paper medical records and the use of fax machines. Yes, some habits die hard, and doctors still use faxes to communicate.
Imagine a world where patient records are digitized and easily accessible. Patients could conveniently book their appointments online and visit any clinic. This would benefit not only those who require care but also practitioners who would find it easier to share information with each other and improve treatment coordination.
The digitization of all processes and the use of technologies already available to the healthcare system are the priority. The system will also have to make a considerable effort if it is ever to accommodate AI within its infrastructure, which is currently, in the words of the CEST, obsolete.
Let's extrapolate: what artificial intelligence technologies are out there that could quickly and efficiently unclog our network?
Top of the list is patient triage in the emergency department, which causes headaches for everyone involved. If it were automated using artificial intelligence, patients could directly indicate their symptoms and reasons for visiting the hospital. An AI model could then determine the nature of the problem and the degree of urgency. AI possesses the collective knowledge of all existing data on all potential diseases, allowing it to surpass the knowledge of a single individual. This process could save enormous amounts of time in the decision-making process and reduce patient waiting times.
CHUM is currently testing the implementation of the first AI model associated with emergency management. The aim is to predict traffic, i.e. the number of people who will come to the emergency department, as well as their profiles and needs. This prediction is made possible using historical attendance data and adding other variables such as the weather or sports calendar. By analyzing these factors, they could detect trends that would not otherwise be recognizable.
The goal of this project is to plan better resources such as beds and the staff on site. By doing so, it would be possible to improve the efficiency of the emergency department by 20%, resulting in better quality treatment for patients seeking medical assistance there.
In the same way as triage in the emergency department, artificial intelligence can be an extremely helpful tool for answering health-related queries remotely. Patients can send photos of their health condition to a model that can identify details in images to receive advice. For instance, the Quebec-based telemedicine platform Dialogue already utilizes AI that simplifies the identification of care that a patient requires based on a description or an image of their symptoms.
Talking of image recognition, some programs are now powerful enough to identify anomalies undetectable to the naked eye in medical images. This technology speeds up the diagnosis of some cancers and improves treatment accuracy. The same applies to the processing of X-rays: an artificial intelligence model can easily analyze them, thus lightening doctors' workloads. Imagia, a company based in Montreal, is already using AI to detect certain types of cancer and develop new personalized treatments.
The above examples are just the tip of the iceberg of what AI can do to improve the healthcare network, and we'd be crazy to miss out.
After all, a hospital operates just like a factory. Both need to manage their production according to demand, some roles need to be automated to help procedures run more smoothly, and they need to analyze their data to perfect their decision-making process. If private companies across all industries are already utilizing advanced technologies like AI and embracing the 4.0 shift, shouldn't we demand the same for our public healthcare system? As taxpayers, we deserve to see our funds invested intelligently and with the highest possible return.
The future of our healthcare system absolutely can only do with AI. Better to start the digital shift as soon as possible! What do you think?
