Scientists have developed a new AI-based method that significantly simplifies the diagnosis of brain tumors in children. The technology relies on “liquid biopsy,” an approach that provides information about the tumor solely through bodily fluids.
The innovative system, named M-PACT, searches for minute genetic traces left by tumor cells in the cerebrospinal fluid. Using a specialized algorithm, the artificial intelligence analyzes chemical changes in the DNA that act like an individual “fingerprint” for each specific tumor.
With this method, the AI can accurately identify the tumor type and its aggressiveness even in cases where the amount of DNA in the sample is extremely small—a barrier that had previously been insurmountable for scientists.
The effectiveness of M-PACT has already been confirmed by large-scale studies, where the method successfully identified 92% of tumors with high precision. Notably, the technology monitors the entire treatment process; the AI can determine how a tumor is responding to therapy without the need for additional surgical intervention.
Furthermore, the system analyzes the “microenvironment” surrounding the tumor and the condition of healthy cells involved in fighting the disease. This revolutionary achievement lays the groundwork for the development of personalized medicine.
Although the method is currently focused on pediatric neuro-oncology, scientists are confident that this innovation will be applicable to the diagnosis of other solid tumors and blood diseases in the future.
The project is the result of international collaboration involving scientists from St. Jude Children’s Research Hospital, alongside researchers from Germany and other leading nations.
