Mayo Clinic researchers and collaborators have shown that artificial intelligence (AI) can analyze routine pathology slides to help classify meningioma, the most common primary brain tumor in adults, and predict a patient’s risk of tumor recurrence.
This study lancet digital healthshow that deep learning models can extract molecular and prognostic information from standard hematoxylin and eosin (H&E) slides, the same type of tissue images already used in daily clinical care. These insights are typically gained through DNA methylation profiling, an advanced genetic test that provides valuable diagnostic and prognostic information, but is expensive, time-consuming, and unavailable in many hospitals.
This is one of many studies that can leverage the strengths of digital pathology by incorporating the past 20 years of genomic and molecular knowledge into AI algorithms. ”
Gheraleh Zadeh, M.D., Chief of Neurosurgery, Mayo Clinic, Rochester, and David C. Pratt and Flora C. Pratt Distinguished Chief Medical Officer, Mayo Clinic Platform
Making advanced oncology insights more accessible
The behavior of meningiomas varies widely. Some are slow-growing and do not grow back after treatment, while others are more aggressive and are more likely to recur. Understanding the risks is critical for patients and their medical teams to decide whether additional treatments, such as radiation therapy, are needed after surgery.
Molecular tests can help identify which tumors are more likely to come back and which respond differently to treatment. However, these tests require specialized techniques and expertise, limiting access for many patients.
Researchers trained the AI to reveal information about tumor biology using tissue samples, pathology images, and clinical data from 672 patients. Leveraging multiple anonymized datasets, including data resources from the Mayo Clinic platform, the model was able to classify meningioma subtypes and predict recurrence risk using standard pathology slides that are already part of routine patient care.
The findings suggest that AI could one day allow clinicians to obtain more detailed tumor information without requiring patients to undergo advanced genetic testing.
Helps guide treatment decisions
For meningioma patients, the risk of recurrence can influence follow-up care, frequency of imaging tests, and whether radiation therapy should be considered. The study found that AI-based predictions remained useful even after accounting for traditional clinical factors such as tumor grade, extent of tumor removal through surgery, and patient age.
The researchers also found that the AI model has the potential to identify patterns of tumor heterogeneity, or differences within the same tumor. This may help explain why some tumors behave more aggressively or respond differently to treatment.
The researchers note that additional prospective studies are needed before AI models can be used routinely in clinical care. Still, the researchers say their findings could lay the foundation for more accessible and personalized care for meningioma patients, and could enable similar AI approaches for other cancers.
“The goal is to make these algorithms easily and easily available around the world to improve patient care in many healthcare settings,” says Dr. Zadeh.
sauce:
Reference magazines:
Landry, AP others. (2026). Deep learning for H&E-based meningioma molecular classification and outcome prediction: A retrospective cohort study. lancet digital health. DOI: 10.1016/j.landig.2026.100986. https://www.thelancet.com/journals/landig/article/PIIS2589-7500(26)00009-9/fulltext
