Improving treatment response prediction and monitoring for tumors in medical imaging
A method to quantify tumor characteristics during treatment using image analysis and machine learning approaches has been developed at McGill University.
By analyzing specific image-based biomarkers during cancer treatment, oncologists monitor disease progression and a patient’s response to therapy. Tumor response to therapy with noninvasive imaging is an essential part of how a physician determines how best to treat the cancer. Unfortunately, current technology only measures one or two dimensions and focuses on tumor size. For irregular tumors or tumors that have infiltrated surrounding tissues, this basic method does not help the doctor accurately predict the proper treatment.
This technology is a computer-implemented method for tracking and visualizing a biomarker to better monitor a patient pre- and post-cancer treatment. By utilizing different medical imaging tools such as CT scans, MRI, or PET, different biomarker calculations will be able to provide a broader view of the tumor. Not only will disease progression be better monitored, but with machine learning approaches, treatment response can be predicted based on a data set of cancer treatments integrated with biomarker analysis.
- Computer-based method using medical tools such as MRI or PET and biomarker analysis to monitor treatment response
- Uses machine learning to predict the likelihood of success for a cancer treatment based on the data sets from biomarker analysis
Filed US, CA