Get in touch with us!
Use our contact form or call us directly.
We will be happy to make you an individual offer and take the time to answer your questions.
Within a 4-week trial phase, you can experience the added value of our software solutions with up to 10 free evaluations.
Would you like to see some sample evaluations in advance? We will be happy to send them to you.
Installation of a secure data connection
To do this, we work with your IT department to install a small piece of software, called a client, which acts as a local PACS node.
It provides an encrypted connection so that a protected interaction between your PACS and our server can take place with maximum efficiency and security.
Collection of suitable image data
To process your image data, we require datasets with the following characteristics:
- T1 DICOM dataset, without contrast, DICOM format: MR Image IOD
- Clear grey-white contrast between grey and white matter and, if possible, hypointense representation of the CSF
- Equipment manufacturers: GE, Philips, Siemens
- Field strength 1.5 or 3 Tesla
- Resolution of the image data in each spatial direction
≤ 2 voxel size ≤ 2.4 mm³
Data protection for your patient data
Data protection conformity is essential.
Therefore, all personal data from the DICOM header are separated from the image data and discarded before transmission.
The technical acquisition parameters, together with the image data and a unique identifier, are pseudonymised and transmitted to the AIRAmed servers for processing via an encrypted connection. Only age and gender are additionally transmitted for comparison with the healthy reference population.
Processing of the data
Our neural network automatically starts processing the data when it is received.
The evaluation of the transmitted data takes only a few minutes.
Results within minutes
We play the evaluation results back into your PACS in the form of a detailed PDF report within a few minutes.
For comprehensible validation of our segmentation results, we play them back co-registered to the original MRI data together with the PDF report. In this way, every automated assignment can be checked for correctness – for complete transparency about the quality of our segmentation.