Engineers at the University of Waterloo have created artificial intelligence (AI) technology that can predict if breast cancer patients may benefit from chemotherapy prior to surgery.
As per The Week, the new artificial intelligence program, which is part of the open-source Cancer-Net initiative directed by Dr. Alexander Wong, could assist unsuitable candidates avoid the severe side effects of chemotherapy and pave the road for improved surgical outcomes for suitable candidates.
Professor of systems design engineering Wong stated, “Determining the proper treatment for a given breast cancer patient is quite challenging at the moment, and it is vital to minimize unwanted side effects from employing medicines that are unlikely to offer genuine benefit for that patient.”
“An AI system that can assist predict whether a patient will respond favorably to a specific treatment equips physicians with the means to prescribe the most effective individualized treatment for a patient, thereby enhancing their chances of recovery and survival,” he continued.
In research directed by Amy Tai, a graduate student in the Vision and Image Processing (VIP) Lab, AI software was trained on breast cancer images created with a new magnetic resonance imaging modality, synthetic correlated diffusion imaging, developed by Wong and his team (CDI).
With knowledge acquired from CDI photos of historical breast cancer cases and information on their outcomes, the AI can predict whether pre-operative chemotherapy treatment would be beneficial for new patients based on their CDI images.
The pre-surgical treatment, known as neoadjuvant chemotherapy, can shrink tumors to make surgery possible or easier and lessen the need for major operations such as mastectomies.
Wong, director of the VIP Lab and Canada Research Chair in Artificial Intelligence and Medical Imaging, stated, “I’m quite excited about this technology since deep-learning AI has the capacity to recognize and discover patterns that indicate whether a patient would benefit from a specific treatment.”
Cancer-Net BCa: Breast Cancer Pathologic Complete Response Prediction Using Volumetric Deep Radiomic Features from Synthetic Correlated Diffusion Imaging was recently presented at the international AI conference NeurIPS 2022.
Through the Cancer-Net project, the new AI system and the whole dataset of breast cancer CDI pictures have been made available to the public so that other researchers can develop the area.