Mass General Brigham has deployed artificial intelligence that reads unstructured clinical data-physician notes, pathology reports, and similar free-text documents-to surface the information that care teams and administrators need. Dr. Alexander "AJ" Blood, a cardiologist at Brigham and Women's Hospital, detailed the system's capabilities in a July 7, 2026 discussion on HIMSS TV.
The technology uses natural language processing to identify diagnoses, medications, lab values, and other relevant details that would otherwise require manual chart review. Dr. Blood said the goal is to cut the time clinicians spend hunting through records and to make that data immediately usable for decision support, quality reporting, and research.
How the AI reads unstructured data
Most electronic health record data sits in structured fields, but a large share of clinically important information lives in narrative notes. Pathologists, for example, write detailed reports that contain nuanced findings. Dr. Blood explained that the AI parses those narratives and converts them into coded, searchable elements. The same approach works for operative notes, discharge summaries, and consultation letters.
The system does not replace the clinician's judgment. It flags and organizes information so that a cardiologist, primary care physician, or case manager can act faster. Mass General Brigham has trained the models on its own data, which Dr. Blood said helps the tool understand the specific language patterns used across its hospitals.
Impact on clinical workflows
Administrative tasks account for a significant portion of a clinician's day. By automatically extracting and structuring information, the AI reduces manual data entry and chart abstraction. Dr. Blood pointed to prior authorization and registry reporting as two areas where the technology already saves hours per week. He said that reducing that manual work gives clinicians more time for direct patient care.
The health system is also using the extracted data to feed analytics dashboards that track population health metrics. Instead of waiting for manual chart reviews, quality improvement teams get near-real-time insights on gaps in care.
Why this matters for healthcare professionals
For clinicians and operations staff, tools that turn unstructured text into actionable data can shrink the documentation burden that drives burnout. The Mass General Brigham approach shows that AI can work within existing EHR workflows, not as a separate system. Professionals who understand how these models are trained and validated will be better positioned to evaluate similar tools at their own organizations. As health systems face tighter margins and workforce shortages, technology that reclaims even a few hours per week per clinician has a direct financial and operational payoff.
Your membership also unlocks:
