CuraMate

CuraMate (formerly known as SATORI) is our cutting-edge toolkit designed for efficient annotation, segmentation, and image feature quantification.

It provides highly customizable annotation forms, intuitive interactive contouring tools, and fully automated segmentation algorithms, allowing for a versatile approach to various annotation tasks. CuraMate aims to guide collaborating users with different roles through their work effectively once a reader study is properly set up and configured, while also ensuring that preventable errors are avoided.

Finally, CuraMate integrates seamlessly with components for training and inference of Deep Learning models. A standout feature is the integrated training loop, enabling users to enhance neural networks continuously by iteratively adding new training data and correcting errors in previous algorithmic outputs.

Downloads:

Flyer CuraMate (PDF)

Flyer Training Loop (PDF)

Versatile Imaging Support

CuraMate handles various medical imaging modalities with standard-compliant import/export.

Seamless Algorithm Integration

Easily incorporates custom or public segmentation algorithms for anatomical or pathological structures.

Advanced Segmentation Tools

Offers state-of-the-art interactive segmentation and correction of relevant structures.

Streamlined Case Processing 

Ensures efficient workflows for case management.

Pharmaceutical Companies & Contract Research Organizations   

• Flexible setup for various image-based clinical studies
• Support for efficient workflows and collaborative reading     
• Compatibility with all common imaging modalities

Clinicians & Researchers

• Comprehensive platform for AI model training and algorithm comparison
• Efficient toolset for annotating and curating image data

Furthermore, our CuraMate training loop benefits both commercial and non-profit research organizations in developing data-driven algorithms, particularly AI segmentation models.

CuraMate

• Highly customizable for an efficient data annotation and curation workflow
• Supports study-specific, usability-optimized workflows for enhanced efficiency
• Integrates diverse image processing tools into a web-based frontend for tailored solutions
• Deployable on private or public cloud infrastructures

Interactive Training Loop

Our interactive training loop accelerates data curation for training segmentation models. It reduces annotation labor costs and enables the use of larger datasets, leading to more robust AI models.

• Optimized image annotation workflow for AI research and development
• Automatic pre-segmentation of new images using preliminary AI models
• Efficient local corrections of false positives/negatives with "scribbles", eliminating the need for full manual annotation
• Uncertainty-based guidance to quickly identify potential errors in large image collections, even without ground truth

CuraMate is continuously evolving with new features developed as extensions, making them available for various projects and customers. Currently, we are focusing on a visual study configurator, enabling users without programming skills to configure a CuraMate instance for new studies effortlessly.