AI Assisted Preoperative Evaluation
PreOpHero is a research prototype developed by ChartHero AI, exploring whether structured AI voice intake, combined with automated record extraction and guideline matching, can meaningfully improve the preoperative anesthesia workflow.
The provider focuses on clinical judgment. The system handles data gathering, organization, and guideline sourcing.
Jie Zhou, MD, MS, MBA, Department of Anesthesiology, Mass General Brigham
Engineering by ChartHero AI
2026 Research Preview
Application Preview
How it works
Three phases, one clinical-ready package.
PreOpHero connects patient data extraction, structured voice intake, and guideline-driven note generation into a single supervised workflow.
Phase 1
Patient data & EHR extraction
Upload clinical documents and enter patient information. The system extracts relevant data and configures the voice agent for that specific patient.
Phase 2
Structured voice intake
An AI voice agent conducts a preoperative interview by phone while the provider can monitor in real time.
Phase 3
Clinical note generation
Data from the call is extracted, matched against clinical guidelines, and structured into a draft note for provider review.
Patient data & EHR extraction
Structured data before the first question is asked.
Before the voice agent makes a call, the system ingests patient records and clinical documents to build a complete picture — so the interview is informed, not cold.
Patient information entry
Demographics, procedure details, scheduled date, and contact information are entered to initialize the encounter.
EHR document upload & extraction
Clinical documents — labs, medication lists, prior notes, imaging reports — are uploaded. The system extracts medications, conditions, and relevant history automatically.
Voice agent configuration
Extracted data is used to customize the voice agent's interview script for this specific patient — skipping questions the chart already answers, and prioritizing follow-up on gaps or concerns.
Structured voice intake
Multi-agent voice intake, clinician-supervised from start to finish.
Each phase of the preoperative interview is handled by a specialized voice agent tailored to that stage's clinical objectives, with optional provider oversight at every stage. Patient-confirmed responses are treated as primary, while chart data informs the conversation but doesn't override what the patient reports.
Phase 01
Identity Verification
The agent confirms the patient's identity before any clinical questions begin, following standard intake verification protocols.
Phase 02
General Pre-Op Intake
Broad data collection covering medications, specialists, conditions, and acute issues — using conversational agents to capture structured data.
Phase 03
Guideline-Driven Follow-Up
Targeted clinical questions adapt based on the patient's condition profile, following preoperative guideline frameworks.
Phase 04
Summary & Confirmation
The agent summarizes everything discussed, confirms with the patient, and closes the call with structured data ready for review.
Optional oversight
Real-time provider oversight
Oversight is available when it helps. Providers can stay hands-off unless they want a closer look or need to step in.
View the live transcript as the conversation happens
Listen in to the call in real time
Nudge the agent's direction if needed
Transfer or end the call at any point
Clinical note generation
From transcript to clinical-ready note.
A multi-step pipeline extracts structured data, matches it against clinical guidelines, and generates a draft note. Every recommendation traces back to the specific transcript excerpt and guideline passage that produced it.
Step 1
Clinical Data Extraction
Structured extraction from call transcript
Step 2
Completeness Validation
Gap detection against required fields
Step 3
Guideline Matching
Clinical guideline & institutional protocol alignment
Step 4
Note Generation
Draft clinical note with citations
Step 5
Provider Review
Final human review and sign-off
Research Team
A collaboration between clinical research and applied AI research and engineering.
PreOpHero was developed through a collaboration between clinical anesthesiology research and applied AI research and engineering.
Clinical Research
Jie Zhou, MD, MS, MBA
Department of Anesthesiology, Mass General Brigham
Led clinical direction, protocol design, guideline selection, and evaluation framework for the preoperative intake system.
Jie Zhou, MD, MS, MBA — Principal Investigator
Applied AI Research & Engineering
ChartHero AI
Led system architecture, voice agent design, clinical data extraction, and LLM-driven guideline matching and note generation — researched, designed, and developed by ChartHero AI.
About ChartHero AI →Interested in the research?
PreOpHero is in active development. If you're a clinician, researcher, or institution interested in preoperative workflow automation, we'd welcome the conversation.
Get in Touch