Discover the POWER of our Data Annotation Services !
Fill out the form below to schedule a personalized demo and see how we can transform your business.
Case Study
Fluoroscopy image annotation for self-expendable valve positioning
Medical Image Annotation Project
RedBrick AI Platform
4 Refugees
in Syria impacted through this project
138945 medical images
annotated with ultra high precision
Annotation Type
Polygon
Classification
Medical Data Annotation
Challenge :
Training AI models to support transcatheter aortic valve implantation (TAVI) requires detailed annotation of anatomical landmarks and medical devices in fluoroscopy images.
The task involved identifying aortic valve cusps, radiopaque valve markers, catheter boundaries, and the lower corners of a self-expanding valve. Their appearance changes with the imaging angle, contrast visibility, and stage of deployment, making consistent annotation challenging.
The project also required clear rules for handling partially visible structures and selecting relevant frames. Only images from the positioning and deployment phases were within scope, with annotations guided by the visibility of each landmark.
Solution :
The project’s annotation approach combined keypoints, polygons, bounding boxes, and frame classification to capture the anatomical and device features relevant to self-expanding valve positioning.
Using RedBrick AI, the workflow followed detailed instructions for identifying visible landmarks, outlining the delivery catheter, and locating the pigtail catheter. Frame classification captured contrast visibility and the presence of a two-cusp view, adding context to the annotations.
This approach was designed to support AI model training and procedural understanding through structured, consistently defined labels.
FEATURES
Implementation:
Protocol-Guided Medical Annotation
The annotation protocol defined the relevant phases of TAVI and the conditions for labeling each structure. Anatomical keypoints identified the visible aortic valve cusps, while device keypoints captured the three radiopaque valve markers and the inferior valve corners as they became distinguishable during deployment. Frames outside the specified procedural phases and valve-in-valve cases were excluded from the annotation scope.
Multi-Type Annotation on RedBrick AI
RedBrick AI supported a combination of annotation tools tailored to each target. Keypoints marked anatomical and valve landmarks, polygons outlined the specified delivery catheter segment, and tight bounding boxes enclosed the visible pigtail section. Frame-level labels recorded contrast visibility and two-cusp views. Where interpolation was used between frames, the protocol required checking that annotations remained correctly aligned in each image.
Structured Review and Feedback
The review workflow allowed reviewers to correct annotations directly or return comments for annotators to address. Task history supported feedback tracking, while issue reporting provided a way to flag ambiguous frames and technical problems. Visibility rules and a dedicated annotation FAQ guided the handling of difficult cases, helping reduce inconsistent interpretations across the team.
Result:
The project scope covered 138945 fluoroscopy images and a four-member team, with an annotation framework encompassing anatomical landmarks, valve features, catheter boundaries, and image-level classifications.
The resulting label structure provides a foundation for training and evaluating AI models focused on self-expanding valve positioning and deployment. It captures both the location of relevant structures and the imaging context needed to interpret them.
Conclusion
This project illustrates the detailed annotation requirements of interventional cardiac imaging. Combining multiple annotation types with procedure-specific instructions creates a structured foundation for medical AI development.
By focusing on visible landmarks, relevant procedural phases, and a defined review workflow, the project supports consistent interpretation of fluoroscopy images used in self-expanding valve positioning and deployment.
How Subul Data Annotation Works For You
Personalized and Fully Managed
Get a dedicated project manager to handle everything from guidelines to quality control, ensuring a seamless experience from start to finish.
Multilayer Quality Control
Experience the assurance of multi-tiered QC processes, including peer review and expert checks, guaranteeing a minimum accuracy of 98%.
Diverse and Scalable Workforce
Access a diverse team of industry experts, assembled within 72 hours, tailored to fit your project's specific needs and scale.
Unbiased and Compliant
Rely on our rigorously trained workforce for unbiased data, perfectly aligned with international privacy and AI standards.
Express Delivery
Benefit from our dynamic workforce, capable of rapidly scaling up to meet tight deadlines and ensure swift project delivery.
Direct Social Impact
Make a difference with every project; our entire workforce comes from refugee and disadvantaged backgrounds, contributing to social goals.
The world’s leading Ai teams Trust Us
