ANNOTATION


Professional data annotation services with our own experts and in-house software – when quality matters!

Data Annotation

Annotation (labeling) of target structures is key to success for AI model performance, with most of the work being data collection and preparation. The quality of the labeling significantly defines the quality of the AI model.

We not only provide high quality annotations to fit the challenge, but also help in planning AI projects, as well as collecting and preparing the data.

Data can be annotated in different ways:

Labeling of entire images, for example with one or more image content classes.

Annotation with region of interests (ROIs) for spatial localization tasks and classification.

Semantic annotation of image content, assigning one or more classes to each pixel.

Our Annotation Services

We provide professional image labeling services that can serve as an essential foundation for training AI models for various applications. In our experience, this can account for up to two-thirds of the project effort.

Our team of annotation experts is ready to relieve you from this time-consuming annotation work.

Applications

 

  • Medical device development
  • Clinical, preclinical & industrial research
  • Clinical trials
  • Quality control

Solutions

 

  • Own software & workflows for high quality annotations
  • Trained annotation team
  • Annotation of large & complex data sets
  • Satisfying legal requirements

Benefits

 

  • Powerful own annotation-software
  • Seamless integration of Development
  • Quality process with integration of external medical experts
  • Close customer collaboration throughout the annotation process
  • Controlling and tracking of project state
  • Keep full control of your data

Researchers in clinical and preclinical healthcare insititutes can also benefit from our data annotation services. Contact us for more information.


Contact Us

Fields of Expertise

Based on successfully completed projects, we offer experience in the following areas:

  • Skeletal and bone annotation in x-ray and CT
  • Dental x-ray annotation (e.g. teeth, nerve channel, facial bones)
  • Soft tissue and organ annotation (e.g. liver, prostate) in CT and MRI
  • Brain tissue annotation in brain MRI
  • Vessel and heart annotation tasks in contrasted CT
  • Marker (fiducials) annotation in x-ray and CT
  • Annotation of implants and tools
  • Cancerous lesions in multi-modal images (e.g. HCC tumors)

 

In general, it's not just about the annotation. You always have to keep an eye on the subsequent application and the AI model to be generated. Therefore we work with interdisciplinary teams of data scientists and medical experts from the very beginning.

Data Scientists


To generate the best possible result from the clients' data, Chimaera provides experienced Data Scientists. They support the selection and preparation of data sets and help define the necessary quality guidelines for the annotations.

Medical & Dental Experts


For medical topics, our customers can outsource the data annotation to our trained team of medical and dental students. Even time-consuming tasks can thus be completed quickly and professionally.

Service Workflow

Since the performance of AI models depends heavily on data annotation, careful quality control is essential, especially for clinical applications.

With our well-designed quality workflow, our customers are fully involved in the annotation process and can involve their own experts in the project whenever desired.

Reference Annotation

Creating reference annotations helps to plan the annotation project:

  • Problem analysis based on test data from the customer
  • Effort estimation for scope and quality of the project
  • Collection of customer requirements for quality specifications

Data Preparation & Collection

Data exchange and preparation needs careful planning and solutions for:

  • Data storage and exchange according to GDPR
  • Anonymization and pseudonymization
  • Data safety and security

Annotation Guidelines

Experience shows that project-specific annotation guidelines are crucial for success:

  • Precise definition of annotation targets and details
  • Specification of tools and process instructions based on reference annotations
  • Continuous adaptation during the project

Project Team

Close interdisciplinary collaboration of:

  • Data scientists and customer engineers
  • In-house annotation experts
  • External medical experts

AI-Assisted Data Annotation

Incremental AI models are created early in the project to support annotation:

  • Continuous AI model updates for improved pre-annotation
  • Development of AI model and annotation go hand in hand
  • Consistent AI result control (reporting)

Controlling

Our controlling measures do not only relate to annotation results, but also to personnel and project management:

  • Regular quality controls
  • Change tracking (documentation)
  • Fine-tuning the annotation process to the task at hand
  • Automated result checks for early error detection
  • Resource and project management

Acceptance and Handover

The top priority in each of our projects is customer satisfaction and transparent delivery of results:

  • Acceptance by the customer and its experts based on the annotation guidelines
  • Support for integration into customer systems
  • No black box solutions but transparent transfer of project results (data and models)

Use Cases

Semi-supervised pixel annotation in surgical endoscopic videos based on DNNs
News

Working on an exciting and challenging M.Sc. thesis at Chimaera as part of the SurgicalAIHubGermany joint research project.



Read more
Comprehensive AI-Workflow at Chimaera
Use Case

Chimaera is here to support you every step of the way on your journey to getting your medical device approved under the MDR. We don't just offer software solutions. We also help you explore the best tech options for your needs and get your product to market quickly and affordably.



Read more
Hackathon Dresden 2024
News

Highly informative days as part of the BMBF project SurgicalAIHubGermany in summery Dresden.



Read more
AI-assisted automation of dental applications
Use Case

Automation is strongly on the rise in the dental industry. With our experience in radiology, we quickly and cost-effectively develop AI solutions in the dental sector as well.



Read more
News

Our team is currently engaged in a fascinating collaboration with Fraunhofer IKTS on a project centered around battery research.



Read more
Use Case

Using the example of the automatic segmentation of hip and femur from computed tomography, this blog post will show how the path to creating an AI model proceeds and what requirements must be met.



Read more
Use Case

In this use case we realized a robust and accurate segmentation of the liver, the spleen, the kidneys and the heart in various Magnetic Resonance Imaging (MRI) sequences.



Read more

Nina

Dr.-Ing. Nina Ebel

Am Weichselgarten 7
91058 Erlangen
Germany

 +49 (0)9131 - 691 388
 +49 (0)9131 - 691 386
ebel(at)chimaera.de