Surgical Intelligence AI

Robotic surgery and Endoscopy image and video labeling at scale to improve model performance

Robotic Surgery and Endoscopy Annotation by iMerit

Surgeon-led teams of iMerit specialize in annotating surgical imagery and video data. With extensive experience in medical-grade video annotation tools, iMerit supports surgical AI use cases, including robot-assisted surgery and other endoscopic and clinical operations. With advanced annotation protocols and guidelines, iMerit ensures high-quality training data for your surgical AI models.

Faster and Efficient Labeling

With hybrid workflows and experts-in-the-loop, iMerit data annotation teams work under the guidance of surgeons, allowing for cost-efficient scaling without compromising quality.

Quick-start Guides and Protocols

With 10+ years of experience in Medical AI, iMerit supports clients with quick-start support, best practices, and labeling protocols across various surgical annotation tools.

Multiple Annotation Types

iMerit Medical AI teams can label images and videos across various annotation types, including polygon, keypoint, bounding box, segmentation, and classification.

Regulatory Compliance

All iMerit workflows leverage an extensive tool ecosystem and can accommodate rapidly evolving HIPAA, SOC-2, CE, and FDA regulatory compliance needs.

Reduce Time to Production

iMerit teams are adept at identifying biases, errors, and disbalances in datasets for faster time to production for surgical AI models. Our consensus workflows help remove annotator biases.

Case Study

Leading Medical Device Manufacturer Works with iMerit to Improve Performance of Surgical AI Model

To enhance patient survival, surgery efficacy, and patient recovery time, the company needed terabytes of surgical videos annotated by experts. With hybrid workflows, the annotations met > 99% accuracy by intersection over union and +/- 3 frame specificity versus gold set data.

After training their model and comparing model performance, the medical device manufacturer found a substantial 12% increase in the recognition accuracy of the model.

Read Case Study

99

Annotation Quality

12

Increase in Model Accuracy

60

Reduction in Annotation Time

"iMerit’s workflows have made them an invaluable partner."

Digital Solutions Engineer, Leading Medical Device Manufacturers

Common Use Cases

Coronary Artery Bypass

iMerit surgeons work with specialized annotators to combine expertise and scalability within a 3-step hybrid workflow. Use cases include surgical analysis and surgical tool classification and segmentation.

Tumor Resection

Using a two-step workflow, iMerit annotators classify and segment tumors. The annotated images undergo review by surgeons to ensure quality while maintaining the speed and lower time-per-task (TPT).

Stomach Ulcer

Surgeons work separately in this consensus workflow to identify ulcers. Once finished, the annotations are merged to mitigate biases and errors and improve the accuracy of the resulting training dataset.