A new development in the field of neuro-oncology could fundamentally change the way brain tumors are diagnosed around the world. Scientists have created an advanced artificial intelligence system that can accurately identify types of tumors of the central nervous system within minutes, replacing the wait that until now could take up to two weeks.

The new technology, called Hetairos, has been developed by researchers at the German Cancer Research Center (DKFZ) in Heidelberg and its results have been published in the prestigious scientific journal Nature Cancer.

The system is designed to analyze the cuttings. tissue microscopic standards, which are used every day in pathology laboratories.

Based only on these digitized images, Hetairos manages to identify more than 100 molecular subtypes of brain and spinal cord tumors with impressive accuracy.

Tumors of the central nervous system are considered among the most complex to diagnose, as they often require advanced molecular analysis to determine their exact nature. Until now, the most reliable standard has been DNA methylation analysis, a process that requires specialized laboratories, expensive equipment and days of waiting.

Hetairos promises to change this reality. To train the system, scientists used more than 11,000 digitized tissue samples from 9,606 patients at 11 medical centers spread across four continents. The results showed that the artificial intelligence can successfully distinguish 102 different tumor subtypes, covering almost the entire current World Health Organization classification of tumors of the central nervous system.

One of the biggest advantages of the system is speed. In clinical trials, traditional molecular diagnosis took an average of 12 days, while Hetairos was able to provide an assessment within just 12 minutes of the sample being digitized.

Including the scanning and material preparation process, results can be available within 24 to 48 hours.

According to the researchers, the system is not limited to just providing a diagnosis. It also shows the level of certainty of the prediction and highlights the specific tissue areas that influenced the algorithm’s decision, enabling doctors to verify and understand the analysis process.

In a direct comparison with experienced specialists, Hetairos was able to distinguish complex morphological patterns that are often difficult to identify even by the most trained neuropathologists.

However, the researchers emphasize that artificial intelligence is not intended to replace the doctors. On the contrary, it is seen as a powerful support tool that can speed up clinical decision-making, help in complicated cases and provide faster access to diagnostics, especially in countries where modern molecular tests are not readily available.

Experts believe that this discovery could mark the beginning of a new era in cancer diagnostics, where artificial intelligence will play an increasingly important role in the detection and treatment of complex diseases.

Source: prizrenpost