Professor Ismail Ben Ayed

The use of artificial intelligence in the medical imaging field

April 23, 2018
Current image acquisition techniques are highly developed. Magnetic resonance imaging, tomography, radiography and sonography have all seen significant advances, and they produce billions of mages every day. Professor Ismail Ben Ayed sees this as an incredible medical potential that has so far been underexploited.



The limits of interpretation

Having access to medical imagery has been of great importance in healthcare for many years. These images allow for a diagnosis to be reached, a therapeutic approach to be defined and the progress of the condition to be monitored. However, medical imaging has always been limited by the capacity of human interpretation. The main obstacle is quantity. Nobody can analyze millions of images. Quality is another challenge, because medical images are often noisy, with poorly defined contours, and therefore, it is difficult for the human eye to achieve a precise reading. As a result, the risk of error remains high.

Artificial intelligence
The solution? Artificial intelligence.

Artificial vision has undergone incredible advances in recent years, thanks in large part to major breakthroughs in optimization and an explosion in the calculating power of computers. Artificial intelligence can process millions of images in only a few seconds, and is able to recognize with extreme precision significant information that would have taken a radiologist months to find.

The power of algorithms
The key to producing these types of results is knowing how to develop highly complex algorithms, but a keen understanding of the medical issue in question is also required. Ismail Ben Ayed works with healthcare professionals, who submit cases to him. He then develops mathematical models that allow for millions of images to be processed, analyzed and interpreted with a high degree of precision.

The challenges to be met in the coming years
In the years to come, experts in the application of artificial intelligence to medical imaging will face significant challenges. The ÉTS Researcher has identified the three main issues:

  1. Compiling precise diagnostic measures automatically or semi- automatically based on medical images.
  2. Predicting events: Computer programs that are capable of processing millions of images may be able to identify those that indicate a risk.
  3. Predicting the results of treatment, such as during a surgical procedure, or monitoring a disease (e.g.: cancer) by measuring the progress of a tumor.

Artificial intelligence is redefining the paradigms and expanding the horizons of medical imaging. Researchers like Ismail Ben Ayed are the masterminds behind this transformation.

See also :
Laboratory of Imaging, Vision and Artificial Intelligence (LIVIA)
Pour information
Sébastien Langevin
Communications Officer
514 396-8427


NouvellesFil RSS