We are the partner which supports you to successfully accomplish Artificial Intelligence (AI) projects. Starting from a project idea we begin with an analysis, to evaluate what technology and solutions are required. We support you gathering the necessary data and annotate it. We design and train the AI models for your product platform while taking into account your hardware requirements. We think through the complete AI project including product integration and product deployment.
AI projects may fail due to missing data, overly complicated annotation workflows, a wrong project setup – or just because you don‘t find a standard approach to solve your problem. Based on our experience in numerous, successfully accomplished machine learning projects, we help you to realize your project!
Plan with us right from the start to find out what you really need:
Our experts assist in data planning, annotation and quality labeling
Our experienced software engineers find a solution to your problem
As a certified ISO 13485 company we support in regulatory affairs
We have a trained annotation team of medical and dental students, a powerful software tool and established processes to ensure quality and consistency. We have experience in setting up large annotation projects for various anatomical structures with a high degree of detail.
The 3D support accelerates the annotation of complex structures and data sets with a high number of layers. Through the powerful combination of AI annotations and our precise brush tools we can perform the annotations efficiently. The innovative quality process functionality guarantees quality-assured annotations for collaboration in larger annotation teams.
You have a project idea? Our Data Scientists will assist you to decide what data you need (amount, quality,...), help you to gather the data and together with our costomer define quality guidelines for the annotation. So you get the most of the data. Our medical experts will annotate the data and control the quality.
For more than ten years Chimaera has experience with Machine Learning algorithms. We are realizing image segmentation, classification and regression tasks with the technology of AI. We offer this technology in a customer-centered workflow and help you to reduce the time to market. We provide customer-based services that combine all stages necessary to incorporate AI into your applications. This comprises data annotation, network design and training on dedicated hardware, adaptations of model design, integration into your existing software as well as technology training to achieve outstanding results and kick-start your developer team to work independently with the new technology.
- Highly trained Machine Learning experts
- Customized AI model design
- Adapt to product requirements and customer interfaces
- In-house training on high performance Cuda servers
- Transpararent source code delivery with
- Trained models
- Training and validation scripts
- Standardized AI backends (APIs)
- Technology transfer and customer training
- Close collaboration throughout the whole process
- Transparent pricing and calculations
- Incremental results and quality reporting
- Lifecycle management
Together with you we develop a scalable deployment strategy while taking into account possible future AI developments. We integrate the trained AI models into your product platform using established deployment platforms as for example Tensor RT or TensorFlow TFX. Models are converted to the final deployment platform and optimized according to a reduced memory footprint and speed. Hardware requirements are captured right from the start to design optimized DNN models which perform optimal on your targeting hardware.
You are planning to create your own AI Team? We can provide you with consulting services regarding necessary tools and hardware or do a kick-off workshop for your complete engineering team.
Our vision is to bring AI solutions and our expertise into your business. In combination with our Engineering Services we offer hands-on training for Deep Learning solutions. At the end of our engineering proects we hand out the source code of the model we have trained for your specific requirements. Additionally we train your team in the new technology. At the end of a project you will be able to apply, refine and retrain your own Deep Learning model.
In this section you find some examplatory use cases we have realized in-house.
In this example we realized a robust and accurate segmentation of the liver, the spleen, the kidneys and the heart in various Magnetic Resonance Imaging (MRI) sequences. The apporach has proven to cope with even non-standardized image intensities and varying contrast in MRI. All aspects of a typical AI service project have been realized in-house by our annotation and software experts. We have achieved an average DICE coefficient across the sequences of 92%.
In this example we have realized the denoising of X-Ray images of the spine. Deep Learning techniques allow to predict the noise of an image. The predicted noise thus can be removed from an image which results in a denoised image.
Medical applications in computer aided diagnosis like fluoroscopy require real-time capable algorithms which can process high frame rates. In the same time there is a growing demand for image processing applications on low-end hardware like mobile devices or embedded solutions. We have developed an algorithm that in a smart way reduces the number of parameters with almost no effect on the accuracy.
We have trained and tested our algorithms with a dataset of 67 images of a seven-class segmentation problem. The target organs are: right and left lung, liver, spleen, right and left kidney. This data set was also used for a two-class-segmentation problem (i.e. differentiating tissue from background). As basis for our compressed algorithm we used the SegNetmodel.
|Number of Parameters||Accuracy|
Reduction of Parameters in %
SegNet1 7-class problem
Our Solution with reduced Parameters
SegNet 2-class problem
Our Solution with reduced Parameters
Applications where AI can be applied to:
Technologies using AI: