Google DeepMind has partnered with the University College London Hospitals NHS Foundation Trust to conduct research on improving radiotherapy.
Before administering radiotherapy, surgeons must go through the segmentation process in which they map the areas where a patient should be treated and which parts of the body should be avoided. Segmentation is tedious and can take up to four hours for patients with head and neck cancers.
Through its collaboration with UCLH, Google DeepMind hopes to make machine learning a key part in speeding up the segmentation process. Google DeepMind will analyze anonymous scans from up to 700 former UCLH patients. The company said it "will treat the patient data [it] [is] using in this project with the utmost care and respect."
Ultimately, Google DeepMind hopes to free up clinicians' time, make radiotherapy planning more efficient and create a radiotherapy segmentation algorithm that can be applied to other parts of the body.