Cloud-based CT data processing for clinical pathology

Purpose and goal: Exciscope builds micro-CT systems to make fast 3D images with high resolution and

contrast of objects with low atomic numbers. In the project, we developed a cloud-based toolchain to perform tomographic reconstructions very quickly by scaling up computational resources when needed. This toolchain is now used by all our customers.

Because we in the project learning about the frameworks that apply to clinical software, the software we developed was developed in such a way that we minimized the amount of work required to incorporate it in a future product. Expected results and effects:

Through the project, we have shortened the time and eliminated risks to develop a medical technology product that meets the requirements for performing virtual histology. Our cloud-based tomographic reconstruction system is also live with all of our existing university customers, giving them the ability to collaborate remotely and utilize advanced algorithms that would not have been as accessible without cloud resources.

Approach and implementation:

The project was carried out through close collaboration with our project partner Mitos. Every two weeks we synchronized our work plans in order to be able to be agile in regards to emerging situations and changes.

The project was divided into 10 work packages, but these were changed during the course when some initial assumptions turned out to be incorrect. With our agile way of working, there were no problems for any of the project parties.