Enhancing Robotic Surgery Safety through Precise Medical Image Annotation

Success Story

Imperial college London



Type of service

Polygon Annotations

Platform used


Services circle icon from humans in the loop with woman at boardi on purple globe background
frames annotated


Model validation circle icon from humans in the loop girl celebrating with robot on purple globe background
conflict-affected individuals

The client

The Multi-Scale Medical Robotics Centre merges medicine and robotics to develop innovative solutions for better patient outcomes. Collaborating with leading universities like ETH Zurich, Imperial College London, and Johns Hopkins University, the Centre focuses on three key research programs:

  • Advanced endoluminal robotic platforms for minimally invasive surgery.
  • Magnetic-guided endoscopic systems for small bowel examinations and treatment.
  • Image-guided robotic interventions for surgical automation.

Together, these initiatives pave the way for transformative advancements in medical robotics.

Researcher Spyridon Souipas at Imperial College London is working on a project which uses AI to detect and localize surgical tools in 3D space during operations. Their goal was to enhance the safety of robotic surgery by identifying surgical tools that could potentially collide with the robot.

The challenge

The customer, a researcher at Imperial College London, faced several challenges in their medical robotics research. One of the main hurdles was the need to detect and localize surgical tools in 3D space during robotic surgeries to enhance safety. Their previous approach using an online labeling tool was inefficient and lacked responsiveness.


I was given the option to work with medical annotators, however my work focuses on surgical tools, and therefore that was not required, although it was great to have the option. Both the annotators and the person I communicated with in the beginning of the project, were extremely helpful and understood my requirements very well.

Spyridon Souipas, Researcher, Imperial College London

The solution

Upon discovering Humans in the Loop through a Google search, the customer chose to collaborate with us due to our prompt response and open communication. They required annotation services for approximately 6000 surgical tool images, with the need for precise polygon annotations to generate the necessary JSON files for network training.

Despite the complexity of the task, our team successfully provided accurate annotations. We drew precise polygons around each surgical tool, specifying the tool present in each image. One of the notable challenges involved the small size of the tools and the intricate surgical background. However, our team overcame these difficulties to meet the customer’s expectations.

The result

Our annotation services impressed the customer with their speed and accuracy, leading to a significant improvement in their network’s performance. They also appreciated our team’s support during administrative processes, which showcased our understanding and professionalism.

5600 images were annotated out of 6000, which is great considering the quality of the images. It saved me about 2 weeks’ worth of time and allowed me to work on other things in the process. The vast majority of the labels were exact, which made my network better at what it does. I had never outsourced annotations before, and I am glad I did this time. I have already recommended Humans in the Loop to my coworkers.

Spyridon Souipas, Researcher, Imperial College London

Interested in implementing a human in the loop in your AI pipeline? Get in touch with us and we would be happy to have a call.