Breast cancer is known to be the leading cause of cancer-related deaths in women globally, and compared to countries in the Middle East, Qatar has one of the highest cancer incidence and death rates. highest breasts.
Regular screening and early detection is crucial, with the American Cancer Society saying that when breast cancer is caught early and is in the localized stage, the five-year relative survival rate is 99%.
There are many breast diagnostic approaches, such as mammography, magnetic resonance imaging (MRI), ultrasound, among others, and the use of artificial intelligence (AI) in these diagnostic technologies becomes more and more popular.
Mammography is the most widely used breast cancer screening tool, but diagnosing cancer from these images is a challenge. One in five breast cancer cases are ignored by radiologists, and according to the American Cancer Society, 50 percent of all women who are screened over a 10-year period will experience a false positive, in which cancer is suspected. wrongly.
A false positive can lead to overtreatment with invasive biopsies and unnecessary stress for patients. A false negative can cause delay in detection and processing.
As digital mammography is used as the standard method for early detection of breast cancer and seems to have limitations, AI comes to the rescue. AI models are developed and used to predict breast cancer in mammograms more accurately than radiologists, thereby reducing false positives and false negatives.
“When you use the naked eye to define abnormalities in image data or when analyzing tissue, you can get the analysis wrong. However, with artificial intelligence, the classification of abnormal or normal tissue is more precise, ”said recently Halima Bensmail, senior scientist and associate professor at the Qatar Computing Research Institute, which is part of Hamad Bin Khalifa University in Qatar. Foundation (QF).
“Due to the large variation in data from patient to patient, traditional learning methods are unreliable and machine learning has evolved in recent years due to its ability to sift through data complex and large in order to be able to detect anomalies, ”she added.
In the field of digital imaging, the quality of mammography images from Qatar is considered adequate, according to a study entitled Breast Cancer Detection in Qatar: Evaluation of Mammography Image Quality Using A Standardized Assessment Tool, funded by the Qatar National Research Fund of QF . But the study also notes that as the country develops additional capacity and awareness for mammography screening, it will be important to continuously monitor image quality.
“People will always wonder how accurate your prediction is, or what is most important when designing a machine learning model: the performance of the model or the accuracy of the model. This answer depends on the application and the domain. But so far we don’t have a machine learning algorithm or artificial intelligence model that gives us 100 percent accuracy of the prediction, ”says Dr Bensmail.
So how does AI work to detect breast cancer or any other type of cancer?
Much of AI relies on machine learning. In machine learning, scientists train the system to learn something very specific, like bad breast tissue versus good breast tissue through pictures.
By training the system with massive amounts of data, it learns to differentiate bad tissue from good tissue. Over time, the algorithm learns to predict with great precision.
According to Dr Bensmail, AI algorithms such as deep learning and neural network-based algorithm provide extremely good results in detecting breast cancer – they provide 90-97% accuracy of the data image, as in mammograms. However, when enough data is not available, machine learning or AI models cannot be built effectively – and this is a challenge in the region.
“In the Arab world, breast cancer is the most frequently diagnosed cancer overall, accounting for around 17.7-19% of all new cancers in 2018, but there is some stigma among Arab women for a breast cancer screening. You really have to motivate and encourage them, which is not always easy. And people are also concerned about data privacy, ”she said.
With AI advancing rapidly, Dr Bensmail predicts that within the next 10 years or so, AI will become even more common in clinical practice. She says disease prediction, especially the classification of breast cancer in the radiology department, is something that happens quickly, especially in the area of image data analysis.
“There are many institutes and groups of researchers who are working on processing image data in a precise way,” she added.