Researchers have developed an artificial intelligence tool that scans medical records to identify cases where breast cancer has spread to other parts of the body, according to Medical Xpress.
Metastatic breast cancer, also called stage 4 breast cancer, occurs when cancer cells travel from the original tumor to other organs such as the lungs, liver, bones, or brain. Identifying when and where breast cancer has metastasized is critical for treatment decisions, but tracking that spread across large patient populations has historically been difficult. Medical records often contain this information buried in unstructured clinical notes written by physicians, which are not easily searchable by standard database tools.
The AI system was designed to read those unstructured notes and flag cases where metastatic disease is documented. Researchers trained the tool on existing patient records so it could learn to recognize the language clinicians use when describing cancer spread. The system then applies that knowledge to new records, identifying patients whose cancer has progressed in ways that might not be captured in structured data fields.
The ability to systematically find these cases has implications beyond individual patient care. Researchers and public health officials rely on accurate data about metastatic cancer rates to study treatment outcomes, allocate resources, and design clinical trials. If metastatic cases are routinely missed in administrative records, the data used to make those decisions may be incomplete.
The tool represents a broader push in oncology to apply AI to problems that involve large volumes of text-based clinical information. Medical records generated by hospitals and clinics contain enormous amounts of detail that structured databases cannot easily capture, and AI systems trained on natural language have shown growing ability to extract meaningful patterns from that information.
Reported by Medical Xpress, the research adds to a body of work exploring how AI can support cancer surveillance and improve the completeness of data used in cancer research and care planning.
