The Evidence Behind Surgical Intelligence
Peer-reviewed. Published. Validated at the institutions that define surgical standards.
Enhancing accuracy of operative reports with automated artificial intelligence analysis of surgical video.
Khanna A, Wolf T, Frank I, et al.
Key finding: AI operative reports: 87.3% accuracy vs 72.8% for surgeon-written reports (p = 0.001). First peer-reviewed head-to-head comparison of AI vs. surgeon documentation in any surgical field.
See how this applies to your system →
Accuracy of warm ischemia time measurement using a surgical intelligence software in partial nephrectomies: A validation study.
Khandekar A, Porto JG, Daher JC, et al.
Key finding: AI-derived WIT within 8.3 sec of ground truth vs. 2.45 min for operative reports (p < 0.001). 100% within 1 min, 97% within 30 sec.
See how this applies to your system →
Defining the standard for surgical video deidentification.
Tollefson MK, Ross CJ.
Key finding: Industry's first standard for surgical video de-identification: at-capture removal of PHI/PII in real time. Automated extracorporeal frame removal at 99.3% accuracy. K-anonymity ≥K=3.
See how this applies to your system →
Surgical intelligence and large language models: leveraging AI for complete and accurate operative reporting.
Khandekar A, Shah H, Freitas P et al.
Key finding: LLM + surgical intelligence found 20.2% of surgical events missing from operative reports and 1.2% discrepant. LLM accuracy: 95.8% overall. Demonstrates AI can systematically audit operative reports against video ground truth at scale.
See how this applies to your system →
Surgical intelligence can lead to higher adoption of best practices in minimally invasive surgery.
Fried GM, Ortenzi M, Dayan D, et al.
Key finding: Critical view of safety (CVS) adoption rose from 33% to 76% in 6 months (p < .001). Cases with full CVS were shorter (44 vs. 57 min, p = .007) and had fewer adverse events.
See how this applies to your system →
Implementing an AI-powered endoscopic surgery video recording system in a large hospital network: lessons learned and future prospects.
Messer N, Nizri E, Lahat G, Szold A.
Key finding: Research at scale using structured surgical video data.
See how this applies to your system →
Routine automated assessment using surgical intelligence reveals substantial time spent outside the patient's body in minimally invasive gynecological surgeries.
Levin I, Bar O, Cohen A, et al.
Key finding: Automated assessment reveals significant extracorporeal time in minimally invasive gynecological procedures. Of 639 total hours, 8.7% were extracorporeal, differing between procedure types (p < .001).
See how this applies to your system →
A novel high accuracy model for automatic surgical workflow recognition using artificial intelligence in laparoscopic totally extraperitoneal inguinal hernia repair (TEP).
Ortenzi M, Rapoport Ferman J, Antolin A, et al.
Key finding: 88.8% overall step recognition accuracy, up to 94.3% per-step in TEP inguinal hernia repair.
See how this applies to your system →
The SAGES Critical View of Safety Challenge: A global benchmark for ai-assisted surgical quality assessment.
Alapatt D, Eckhoff J, Lyu Z, et al.
Key finding: Theator's surgical intelligence validated for critical view of safety detection across multiple institutions.
See how this applies to your system →
The film room: using artificial intelligence to facilitate video review for urology trainees.
Henning GM, Findlay BL, Cohen TD, et al.
Key finding: AI-facilitated video review improves surgical training efficiency and self-directed learning for urology trainees.
See how this applies to your system →
A step toward modernization of urologic training: Incorporation of a novel surgical intelligence platform for robotic prostatectomy video review.
Henning GM, Findlay BL, Cohen T, et al.
Key finding: 94% of trainees agreed or strongly agreed video review helped develop surgical skills. Largest published study of urology trainee video review.
See how this applies to your system →Validity of video-based general and procedure-specific self-assessment tools for surgical trainees in laparoscopic cholecystectomy.
Balvardi S, Semsar-Kazerooni K, Kaneva P, et al.
Key finding: Automated video capture enabled assessment and validation of several self-assessment tools for surgical trainees.
See how this applies to your system →
Situating artificial intelligence in surgery: a focus on disease severity.
Korndorffer JR Jr, Hawn MT, Spain DA, et al.
Key finding: Foundational positioning paper establishing the clinical case for AI in the operating room — safety, efficiency, and quality as primary drivers.
See how this applies to your system →Artificial intelligence in surgery: promises and perils.
Hashimoto DA, Rosman G, Rus D, Meireles OR.
Key finding: Foundational framing of AI's role in surgery — opportunities, risks, and the path to clinical adoption.
See how this applies to your system →
Accurate detection of out of body segments in surgical video using semi-supervised learning.
Zohar M, Bar O, Neimark D, et al.
Key finding: 99.3% accuracy in automated detection and removal of extracorporeal video segments using semi-supervised learning.
See how this applies to your system →
Video transformer network.
Neimark D, Bar O, Zohar M, Asselmann D.
Key finding: Novel transformer-based architecture for surgical video understanding outperforms prior CNN-based approaches.
See how this applies to your system →"Train one, Classify one, Teach one" - Cross-surgery transfer learning for surgical step recognition.
Neimark D, Bar O, Zohar M, Hager GD, Asselmann D.
Key finding: Transfer learning enables step recognition across surgical specialties, dramatically reducing data requirements for new procedure types.
See how this applies to your system →
Impact of data on generalization of AI for surgical intelligence applications.
Bar O, Neimark D, Zohar M, et al.
Key finding: Systematic analysis of how dataset scale and diversity drive AI generalization across surgical procedure types.
See how this applies to your system →Association of surgical resident wellness with medical errors and patient outcomes.
Hewitt DB, Ellis RJ, Chung JW, et al.
Key finding: Lower wellness associated with 53% increase in odds of reporting a major medical error.
See how this applies to your system →
Estimating the attributable cost of physician burnout in the United States.
Han S, Shanafelt TD, Sinsky CA, et al.
Key finding: Annual burnout costs approximately $7,600 per physician. US healthcare system-wide cost estimated at $4.6 billion annually in turnover and lost productivity.
See how this applies to your system →
The business case for investing in physician well-being.
Shanafelt T, Goh J, Sinsky C.
Key finding: Each physician replaced costs at least $500K–$1M.
See how this applies to your system →Documenting or operating: where is time spent in general surgery residency?
Cox ML, Farjat AE, Risoli TJ, et al.
Key finding: Surgical residents spend at least 30% of their time on EHR tasks. One third of EHR usage by interns occurred outside scheduled 12-hour shifts.
See how this applies to your system →A video is worth a thousand operative notes.
Dimick JB, Scott JW.
Key finding: Foundational editorial establishing that surgical video is a superior record of intraoperative events compared to narrative operative reports.
See how this applies to your system →
Introducing surgical intelligence in gynecology: Automated identification of key steps in hysterectomy.
Levin I, Rapoport Ferman J, Bar O, et al.
Key finding: Theator model achieves high accuracy for automated step identification in hysterectomy.
See how this applies to your system →
Automated surgical step recognition in transurethral bladder tumor resection using artificial intelligence: transfer learning across surgical modalities.
Deol ES, Tollefson MK, Antolin A, et al.
Key finding: Fully automated computer vision algorithm for high-accuracy annotation of TURBT surgical videos.
See how this applies to your system →
Automated identification of key steps in robotic-assisted radical prostatectomy using artificial intelligence.
Khanna A, Antolin A, Bar O, et al.
Key finding: 92.8% concordance between artificial intelligence‒enabled automated video analysis and manual human video annotation.
See how this applies to your system →Intellectual Property
30+
Granted Patents
Across 7 countries: United States, Israel, Europe, China, Republic of Korea, Japan, and Australia. A deep, defensible technical moat spanning the full surgical intelligence stack.
Automated operative report generation from video
US 10,943,682; US 11,769,207; US 12,334,200; EU EP3928325
Surgical video analysis and event detection
US 10,878,966; US 11,763,923; CN 202080029504.X; KR 10-2572006; AU 2020224128; JP 7596269
Surgical competency assessment
US 11,348,682
Decision support and predictive outcomes
US 10,886,015; US 11,452,576; US 12,315,609