Different patients with the same type of cancer can have different responses to the same medication, which leaves doctors in a tough spot: how do they know which treatment will have the best response? If they get it right, their patient may enter remission; but if they're wrong the patient's health will deteriorate. Now researchers at Western University might have the answer. They developed machine learning algorithms – a branch of artificial intelligence – that crunch genetic data to determine the most likely treatment response and allow more personalized treatment regimens.
"Artificial intelligence is a powerful tool for predicting drug outcomes because it looks at the sum of all the interacting genes," said lead researcher Peter Rogan. "The earlier we treat a patient with the most effective medication, the more likely we can effectively treat or possibly even cure that patient."
The researchers used a set of 40 genes that are found in 90 per cent of breast cancer tumors for their analysis of data from cell lines and tumor tissue samples from around 350 cancer patients who were treated with at least one of the two chemotherapy drugs paclitaxel and gemcitabine.
They then set their computers to work crunching the data and identifying associations between the drug and patient genes. Their machine learning tool managed to predict gemcitabine resistance and paclitaxel sensitivity with 84 per cent accuracy, paclitaxel resistance with 82 per cent accuracy, and gemcitabine response (i.e. remission or not) with 62 to 71 per cent accuracy.
The researchers now plan to refine their algorithms and feed the system more data to improve the predictions.
No comments