HaCancer is a disease that afflicts millions of patients worldwide. Most cases are treatable, but many others are fatal. One of the main factors in being able to act in time is precociousness of diagnosis. Because in that way, medicine can be used to attack and cure disease.
A diagnostic system based on artificial intelligence has been developed. We owe this achievement to a team of neurosurgeons and engineers from several US universities.
Their work is based on analyzing tumor samples taken during surgery and using high-speed imaging, which can detect genetic mutations very quickly. In fact, a diagnosis can be made within 90 seconds.
Brain tumors are the cornerstone of research
A team of scientists and doctors conducted a study on a sample of more than 150 patients suffering from diffuse glioma, the most common brain tumor and also the deadliest.
This new system allowed them to identify mutations used by the WHO (World Health Organization) to define molecular groups of diseases, with a diagnostic accuracy of over 90%.
This diagnostic advance is a major advance for medicine. Because the earlier it can be detected, the better the outcome of treatment. One of his leaders in the program, his Dr. Todd Holona neurosurgeon at the University of Michigan.
“This AI-based tool has the potential to improve access and speed to diagnosis and treatment for patients with life-threatening brain tumors.” Holon I got it.
“DeepGlioma paves the way for accurate and more timely identification, giving healthcare providers a better chance to define treatments and predict patient outcomes.”
This new diagnostic system is called “DeepGlioma”. Previously, surgeons had neither the ability nor the way to know how to distinguish between different types of diffuse glioma during surgery.
The new method was conceived in 2019 and began testing by combining an optical imaging system with deep neural networks to take instant snapshots of brain tumor tissue.
“DeepGlioma paves the way for accurate and more timely identification that gives healthcare providers a better chance to define treatments and predict patient outcomes,” added Hollon. .