Skip to main content
Blog post

How did Concordia help enhance health care through artificial intelligence in 2021?

November 25, 2021
By Mojtaba Hasannezhad

Credit: CIO

Artificial intelligence (AI) uses computer science to design machines capable of learning, reasoning, establishing communication, and taking decisions, all considered features typical of humans. Many industries now employ AI to enhance precision and accuracy, performance, time management and efficiency, and cost savings. AI is gaining popularity in the health care industry, and altering the atmosphere of the related research.

Health care researchers at Concordia are attempting to integrate AI in their studies through technological innovations.. Below is some of the remarkable AI work done by health care studies researchers at Concordia in 2021.

Cervical cancer

As one of the most common types of cancer among the female population, cervix cancer affects over half a million women annually and leads to 300,000 deaths globally. Nevertheless, it can be effectively prevented. Treatment depends on how severe the condition is and whether necessary resources are available at the time of diagnosis. Recent research indicates that using sophisticated imaging technology can result in better treatment planning. 3D Ultrasound imaging is considered the most suitable means of screening cervix cancer as it is economical, non-invasive, easy to use, and can be done in real-time. Concordia researchers have developed novel AI methods to improve screening accuracy to provide helpful guidance for the delivery of the correct dose of radiation to the target tissue during radiotherapy therefore enhancing the treatment.

Breast cancer

Breast cancer is a principal cause of mortality among females around the world. The imaging modality most commonly used to screen this type of cancer is mammography, which unfortunately has certain deficiencies. Mammography is not good at differentiating between  dense tissues and cancerous ones. Concordia scientists have devised AI-based methods to correctly classify both benign and malignant ultrasound images. Such methods can diagnose suspicious breast nodules and help detect the beginning of cancer at a rate of 99.09%.

Lung cancer

Lung cancer is the primary cause of death from cancer globally, which is characterized by different microscopic anatomy types. Lung Adenocarcinoma (LAUC) is one example that has recently been reported as the most common. If the invasiveness of lung nodules is timely and accurately identified, then proper treatment planning can be done. The main imaging method currently used to assess the invasiveness of LAUCs is chest CT. However, the problem is the subjectivity and low accuracy of the results. Researchers at Concordia have established an AI-based framework to accurately predict and classify LAUCs. The findings show an accuracy rate of 87.73% in LAUCs classification.

Neurological movement disorders

A sharp increase in the seniors’ population by 2050 means a rise in the number of people influenced by neurological movement disorders. Pathological hand tremor (PHT) is a typical symptom of Parkinson’s disease (PD) and essential tremor (ET), influencing such functions as motor manual targeting and movement. Successful treatment of the symptoms depends on the accurate and timely diagnosis. However, as the corresponding symptoms have overlapping features, specialized diagnostic methods are needed to accurately differentiate PD from ET. They share a number of common symptoms, one of which is PHT, affecting functions such as targeting, coordination, and voluntary movements. Researchers at Concordia have developed a deep-learning-based model assessing the kinematics of the hand and correctly classifying patients into PD or ET. The model ensures a great differential diagnosis accuracy level of 95.55%.


COVID-19 has considerably changed the world as we know it. It is a highly infectious disease that is rapidly spreading worldwide, which makes early diagnosis very important. Common symptoms include cough, fever, and fatigue. It has resulted in high ICU admissions, making it necessary to reach fast and viable diagnosis. Diagnosis of COVID-19 in its early stages helps health care experts and officials control the transition chain, resulting in a decline of positive cases. Tools such as CT scans and X-ray images provide specific symptoms connected with this illness. However, what makes diagnosis difficult is its overlap with other lung infections such as pneumonia and lung cancer. Thus, there has been a great interest among scientists to find diagnosis solutions based on deep learning to help identify positive cases. Accordingly, research conducted at Concordia has proposed a novel AI-based framework using a database of X-ray images for COVID-19 diagnosis, which has an accuracy level of 95.7%.

About the author

Photo of Mojtaba Hasannezhad

Mojtaba Hasannezhad is a PhD candidate in Electrical and Computer Engineering. He received his MSc in Electrical Engineering from Tarbiat Modares University, Iran; his research was focused on telecommunication circuit and system design.

He is currently working on artificial intelligence-based living assistants that can serve humans in many respects to facilitate life and improve its quality. These devices have many applications, especially in healthcare automation. Mojtaba's research is in cooperation with industrial sponsors Microchip Canada, Ottawa, and Moonshot Health, Montreal. His interest is machine learning, human-machine interfaces, smart assistants, and speech processing.

Back to top

© Concordia University