When surgeons approach the edge of a brain tumor with their scalpels, they confront a difficult choice: to cut away portions of healthy brain tissue to guarantee complete tumor removal, or to opt for a more cautious approach, potentially leaving some of the harmful cells behind.
Researchers in the Netherlands have announced a breakthrough in utilizing artificial intelligence to equip surgeons with critical information about tumors, aiding them in making this crucial decision.
Described in a study published in the journal Nature, this technique involves a computer analyzing portions of a tumor’s DNA, discovering specific chemical alterations that can provide a detailed diagnosis of both the type and even subtype of the brain tumor.
How will this diagnosis help Surgeons?
This diagnosis is generated in the initial phases of surgery and spans several hours, it aids surgeons in determining the level of aggression required to operate, according to the researchers. In the future, this method may also guide doctors toward treatments customized for a specific subtype of tumor.
Jeroen de Ridder, an associate professor in the Center for Molecular Medicine at UMC Utrecht, a Dutch hospital, and one of the study's leaders, highlighted the crucial importance of knowing the tumor subtype during surgery. He featured the unique capability they have now established, enabling this highly precise, detailed diagnosis to be conducted right in the midst of the surgical procedure.
How was it tested?
The deep learning system, called Sturgeon, was initially tested on frozen tumor samples from previous brain cancer surgeries. Impressively, it provided accurate diagnoses for 45 out of 50 cases within just 40 minutes of starting genetic sequencing. In the remaining five cases, it refrained from delivering a diagnosis due to unclear information.
After this, the system underwent testing in 25 live brain surgeries, with the majority involving children. This was done in co-existence with the conventional method of examining tumor samples using a microscope.
The innovative approach successfully delivered 18 correct diagnoses, while in the remaining seven instances, it didn’t meet the required confidence threshold. It turned around its diagnosis in under 90 minutes, a duration conducive to informing decisions mid-operation, as per the study.
Currently, doctors rely on both microscopic examinations and, in some cases, more extensive genetic sequencing for analyzing brain tumor samples.
However, not all hospitals have access to this advanced technology and even for the ones equipped with it, obtaining results can be a prolonged process, taking several weeks, according to Dr. Alan Cohen, the director of the Johns Hopkins Division of Pediatric Neurosurgery and a specialist in cancer treatment.