A tool developed in Barcelona, ​​based on artificial intelligence, will improve the diagnosis of brain tumours. Researchers are thinking about the goal of reaching a diagnosis without having to resort so often to the currently almost inevitable biopsy, which in the case of the brain represents a complicated surgical intervention for professionals and disturbing for patients.

The diagnosis of brain tumors is based on the evaluation of magnetic resonance images before and after administering contrast (substances that allow to improve visualization or identify possible lesions during the scan), although most of the time with the radiological test is not enough to avoid subjecting the patient to a diagnostic neurosurgical procedure.

Discern, as the AI ​​designed by the radiomics group of the Vall d’Hebron Institute of Oncology (VHIO), in collaboration with researchers at the Bellvitge hospital, is called, has learned to differentiate with a probability of 78% success rate among the three most common types of brain tumors. Glioblastoma multiforme, brain metastases from solid tumors and primary lymphoma of the nervous system account for 70% of malignant cancers and require different therapeutic approaches, which is why it is essential to identify them precisely and unambiguously.

The researchers have taught the AI ​​what are the characteristics of tumors that appear in MRIs, starting from 50,000 voxels (a voxel is the smallest unit that can be processed in 3D MRI images, equivalent to a pixel in a 2D object) of 40 patients already diagnosed. The tool has been validated in more than 500 additional cases, and the machine is able to correctly find the repeating patterns in each of the tumors in 78% of the tests, a higher proportion than the methods obtained current

“We asked the tool for a final diagnosis, but it also gives you a diagnosis for each of the voxels in the image, so when it provides the same diagnosis for 80% of the voxels the doctor can be more confident about the diagnosis,” he explains Raquel Pérez-López, head of the VHIO radiomics group. What happens in the cases (22%) of no success? “In these cases, Discern gives the voxels diagnoses randomly, so if this happens the doctor can see that the machine is not clear and reject the diagnosis it gives,” the radiologist explains.

In short, the tool deeply analyzes the image data that the human eye is not able to perceive and finds the complex patterns that characterize a tumor compared to others. According to Pérez-López, the 78% success rate can be improved: “Experience in other projects indicates that, when you train the machine with more cases, it improves”. At the same time, the researchers are studying the idea of ​​expanding the possible selection to the rest of the brain tumors.

The results of the study have been published in the scientific journal Cell Reports Medicine and the authors have developed open access software so that the tool can be used in any hospital to continue improving the diagnostic system.

The application of this advance will not be immediate. “New biomarkers must go through a series of processes. A prospective clinical trial must be carried out and these trials are costly in terms of time and money”, explains the radiologist. In his opinion, five years is a realistic deadline for Discern, after clinical trials that will last no less than two years, to be available.