A machine-learning-based “digital twin” was able to predict brain tumor responses in individuals with glioma, according to a study published Jan. 6 in Cell Metabolism.
Led by a team from Ann Arbor-based University of Michigan, researchers developed the digital brain twin using patient data obtained through blood draws, metabolic measurements of the tumor tissue and the tumor’s genetic profile, according to a Jan. 12 news release from the university.
Here are three things to know from the study:
- The digital twin calculated how quickly or slowly a tumor’s cancer cells consumed and processed nutrients with “high accuracy,” the release said.
- Mapping the metabolic activity helped determine which treatment strategies would be effective for individual tumors.
- Researchers also tested the digital twins on how a tumor would respond to the drug mycophenolate mofetil. The twins correctly identified which tumors could bypass the drug’s effects.
Read the full study here.
Leave a Reply