Scieists model a new artificial ielligence to help Treatme of brain cancer have developed that during surgery it can include other resectable parts of the tumor 10 seconds to diagnose
Technology developed by scieists from the University of Michigan and the University of California FastGlioma It’s called and they say it uses conveional methods to ideify residual tumor parts during surgery. Todd Hallon, a neurosurgeon at the University of Michigan and the principal investigator of the study, says:
“FastGlioma is an artificial ielligence-based diagnostic system that has the poteial to change the field of neurosurgery by immediately improving the comprehensive manageme of paties with diffuse glioma (a type of brain cancer). “This technology works faster and more accurately than the curre standard methods in terms of tumor diagnosis and can be used in the diagnosis of other brain tumors in children and adults.”
Accuracy of FastGlioma in brain tumor diagnosis
To develop this AI-based model, neurosurgery teams analyzed fresh, unprocessed samples from 220 glioma paties. In their study, published in the journal Nature, they claimed that FastGlioma was able to detect and quaify residual tumors with an average accuracy of approximately 92%.
Also about the performance of FastGlioma, it is claimed that this artificial ielligence technology failed to detect high-risk tumors only 3.8% of the time, while this rate is approximately 25% for curre routine methods.
FastGlioma uses microscopic optical imaging systems combined with artificial ielligence developed with training data to detect residual brain tumor.
Ultimately, the scieists claim that the technology, in addition to being an accessible and cost-effective tool for neurosurgery teams working on glioma, can also accurately detect residual fragmes in non-glioma tumors, including pediatric brain tumors such as medulloblastoma. They also plan to focus on detecting other cancers in the future, including lung, prostate and breast cancer with FastGlioma.
