Machine Learning is used to Predict How a Brain Tumor will Grow

Glioblastoma multiforme (GBM) is a type of Brain Tumor where most people only live for one year after being diagnosed. It is hard to treat because its core is very dense, it grows quickly, and it is in the brain. Estimating how these tumours spread and how fast they grow is helpful for doctors, but it is hard to get this information quickly and accurately for each patient.

Researchers from the University of Waterloo, the University of Toronto, and St. Michael’s Hospital in Toronto are working together to look at MRI data from more than one person with GBM. Machine learning is being used to fully analyse a patient’s Brain Tumour so that they can better predict how cancer will spread.

Researchers looked at two sets of MRIs from each of the five GBM patients who wished to remain anonymous. The patients went through a lot of MRIs, waited a few months, and then went through another set of MRIs. Because, for unknown reasons, these patients chose not to get any treatment or medical help during this time, their MRIs gave scientists a unique chance to see how GBM grows when it isn’t stopped.

The researchers used a deep learning model to turn the MRI data into estimates of the parameters that are unique to each patient. These estimates feed into a model that predicts how a GBM will grow. This method was used on both real and fake Tumours from patients, whose true characteristics were known, so the model could be tested.

Cameron Meaney, a PhD candidate in Applied Mathematics and the study’s lead researcher, said, “We would have loved to do this analysis on a huge data set.”

However, because of the nature of the sickness, it is difficult to achieve this goal, as patients typically begin treatment as soon as they become aware of their condition. That’s why the chance to compare five tumours that hadn’t been treated was so rare – and important.

Now that scientists have a good model of how GBM grows when it is not treated, they need to add in how treatment affects the Brain Tumor. Then there would be thousands of MRIs in the data set instead of just a few.

Meaney says that getting access to MRI data and having mathematicians and clinicians work together can have huge effects on patients in the future.

Read More: 6 ways SEO Specialists are now using ChatGPT

Stay Connected!

Are You Looking For Machine Learning?