Even before a treatment is determined, brain tumors present a complicated problem for doctors, but a new AI-based procedure may be ready to help.
As a recent article from Dallas-based UT Southwestern Medical Center describes the dilemma, doctors preparing to treat gliomas, a type of tumor that starts in the glial cells of the brain or spine and comprises roughly 30 percent of all brain tumors, often have to obtain tumor tissue for analysis by surgery that “can sometimes be time consuming and risky – particularly if tumors are difficult to access.”
A new study by UT Southwestern, however, shows describes a new AI-powered tool that can identify a specific genetic mutation in a glioma tumor – with more than 97 percent accuracy – simply by examining 3D images of the brain. While scientists have recently testing an array of imagine techniques, this technology could potentially eliminate the need for those pretreatment surgeries.
"Knowing a particular mutation status in gliomas is important in determining prognosis and treatment strategies," said Joseph Maldjian, M.D., chief of neuroradiology at UT Southwestern’s O’Donnell Brain Institute. “The ability to determine this status using just conventional imaging and AI is a great leap forward.”
According to the article, this study distinguishes itself from other, similar research efforts in its level of accuracy and its efficiency. Specifically, “mutation status was determined by analyzing only a single series of MR images, as opposed to multiple image types,” and only “a single algorithm was required to assess the IDH mutation status in the tumors. Other techniques have required either hand-drawn regions of interest or additional deep-learning models to first identify the boundaries of the tumor then detect potential mutations.”
Said Dr. Maldjian, ““We’ve removed additional pre-processing steps and created an ideal scenario for easily transitioning this into clinical care by using images that are routinely acquired,” adding that similar methods may be used to identify other important molecular markers for various cancers.
To improve the process of detecting enzyme mutations and deciding on appropriate therapy, Maldjian’s team “developed two deep-learning networks that analyzed imaging data from a publicly available database of more than 200 brain cancer patients from across the U.S.”
Next, Maldjian explained, the team will test the deep-learning model on larger datasets for additional validation before deciding whether to incorporate the technique into clinical care.
“In the big picture, we may be able to treat some gliomas without operating on a patient,” Maldjian said. “The field of radio-genomics is exploding with possibilities.”