University of Waterloo engineers have created an AI tech that can predict if women with breast cancer may benefit from chemotherapy before surgery.
New AI algorithm to help people avert sides effects of chemo
The latest AI algorithm, which is part of the open-source Cancer-Net initiative spearheaded by systems design engineer Dr. Alexander Wong may help unsuitable candidates to avert any serious side effects of chemo and create a way for better surgical outcomes for suitable individuals. Wong is the VIP Lp director and the Canada Research Chairperson in Artificial Intelligence and Medical Imaging.
Wong said that choosing the best course of action for a specific woman with breast cancer is currently highly challenging, and it is critical to prevent unwanted side effects from being caused by therapies that are not likely to actually benefit the individual. Doctors can better prescribe the most individualized therapy for a person to increase recuperation and survival with the help of an AI model that can determine whether a patient is most likely to respond favorably to a particular treatment.
In an initiative led by Amy Tai, In a project directed by Amy Tai, a grad student in the Vision and Image Processing (VIP) Lab, artificial intelligence (AI) software was programmed using pictures of breast cancer taken using a novel magnetic resonance imaging (MRI) modality termed synthetic correlated diffusion imaging, developed by Wong and his group.
AI could show if pre-operative chemotherapy is beneficial
The AI could determine if pre-operative chemo would be advantageous for patients depending on CDI images using data collected from CDI photos of old cases of breast cancer and data on their results.
Popular as neoadjuvant chemo, the pre-surgical treatment shrinks tumors making surgery easier or possible and can reduce the need for surgery like mastectomies. Wong explained that they are optimistic about the tech since deep-earning AI could potentially help discover patterns relating to whether a patient can benefit from a specific treatment.
The AI algorithm and the breast cancer CDI images dataset are available publicly through the Cancer-Net initiative to help advance research in the field.