Skip to main content

Operative Reports Generated from What Actually Happened.

Surgery-to-Text® analyzes intraoperative video from your procedures, identifies every step, landmark, and event, and delivers a billing-ready operative report into your EHR before you scrub out.

No dictation. No templates. No memory. The video is the source.

Request a Demo

The Operative Report Is Broken.

The operative report drives coding, billing, quality, medico-legal defense, and care decisions. And yet:

∼73% accuracy

Inaccurate surgeon-written reports.

Khanna et al., JACS 20251.

∼30% undocumented

Procedures with undocumented elements.

AHIMA; Khandekar et al., BJUI 20243.

69% burnout

Surgeon burnout from documentation.

Medscape 2024.

$118/denial rework

Costly claim denial reworks.

AAHAM benchmarks.

Written from memory. Hours later. Generic templates.
Surgery-to-Text® eliminates the root cause.

How It Works

Capture
Step 1

Capture

Existing laparoscopic or robotic camera feed. No new hardware. No workflow change. Any manufacturer. Automatic capture, de-identification, secure upload.

Understand
Step 2

Understand

Proprietary vision-language model trained on 600,000+ procedures, 9 specialties, 180+ types. Identifies steps, instruments, anatomy, milestones, events, complexity. Not frame detection. Contextual understanding of the procedure's narrative arc.

Generate
Step 3

Generate

Procedural understanding + EHR patient context (HL7/FHIR) = complete, structured operative report. Findings, technique, instruments, complications, CPT/ICD specificity. <30 seconds.

Deliver
Step 4

Deliver

Directly into Epic or Oracle Health. No portal. Surgeon reviews, edits, finalizes.

Surgery-to-Text demo

In Surgery, Actions Speak Louder Than Words.

Voice tools transcribe what clinicians say. In clinic, that works. Surgery is different. Hands occupied, no narration. The clinical reality is on screen, inside the body cavity, not in ambient audio.

Surgery-to-Text® analyzes what is done, not what is said.

Peer-Reviewed. Published. Validated.

Metric
Result
Source
AI vs. manual accuracy
87.3% vs. 72.8% (p = 0.001)
Khanna et al., JACS 20251. Mayo. n=158.
Current accuracy
>95%
Theator platform metric.
Events missing from reports
20.2% of 2,909 events
Khandekar et al., AUA 20254. UHealth Miami. n=180.
Documentation time reduction
∼70%
UHealth Miami pilot data2.
Report generation time
<30 seconds
Khandekar et al., BJUI 20243. UHealth Miami. n=61.
Documentation ROI
At least +$186/case
Theator deployment data.

9 specialties and counting. 180+ procedure types. Robotic prostatectomy to laparoscopic cholecystectomy. Growing continuously.

Request a Demo to See It on Your Cases

The Most Accurate Operative Report in Surgery. In Seconds.

Request a Demo to See It on Your Cases
THEATOR

1Khanna A, Wolf T, Frank I, et al. Enhancing accuracy of operative reports with automated artificial intelligence analysis of surgical video. J Am Coll Surg. 2025;240(5):739-746.

2University of Miami Department of Urology: Epic-integrated Surgery-to-Text® pilot demonstrating 70% reduction in time to final op note, median 6 edits per case.

3Khandekar A, Porto JG, Daher JC, et al. Accuracy of warm ischemia time measurement using a surgical intelligence software in partial nephrectomies: A validation study. BJUI Compass. 2024;5(12):1263-1268.

4Khandekar A, Shah H, Freitas P et al. Surgical intelligence and large language models: leveraging AI for complete and accurate operative reporting. American Urological Association (AUA) Annual Meeting 2025.