Machine Learning Identifies Cardiotoxicity Risks in Breast Cancer Patients • Healthcare-in-Europe.com

Machine Learning


Portrait photo of Dr. Paaladinesh Thaveniranathan

Dr. Paaladinesh Thavendiranathan

Image courtesy of Dr. Thavendiranathan

This study addressed the challenge of predicting cancer therapy-related cardiac dysfunction (CTRCD). This is “still a challenge,” the researchers said. Clinical risk models and traditional cardiac magnetic resonance (CMR) analyses are “limited” in predicting HER2+ targeted therapy (HER2-TT) CTRCD risk, according to a research paper published in the European Heart Journal Supplements.

Dr. Thavendiranathan, a professor of medicine at the University of Toronto in Canada, explained that current techniques for risk stratification are “not very effective.”

Promising research performance

The Canadian study aims to determine whether deep learning (DL) can better predict CTRCD than clinical risk scores or traditional quantified imaging measurements using CMR cine imaging during or early cancer therapy.

Three prospective studies included women with early HER2+ breast cancer receiving serial anthracyclines and trastuzumab. To create various machine learning models to predict CTRCD, patients were seen continuously during treatment with repeated cardiac imaging (echocardiography and CMR), before and after anthracycline.

“This model performed better than clinical risk factors, echocardiographic measurements, traditional cardiac MRI measurements or biomarkers,” said Dr. Thavendanathan. “When we validated our findings in a different cohort, the model continued to work well.”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *