Interactive image processing of Big Data in medical diagnostics

Purpose and goal

The overarching goal of the project was ´To elicit new knowledge to achieve groundbreaking increase of performance for large-scale image analysis in medical diagnostics´. This knowledge creation goal has been well achieved. The main technical design decisions have been successful since the prototype has shown to combine powerful analysis with interaction capacity, an idea that was not verified at the project start. Preliminary validation by end users have confirmed the potential for future usefulness.

Expected results and effects

The main result is the developed prototype, by which the project´s solution ideas for large-scale image analysis has been tested and found to be valid. The created knowledge will, e g, be put to industrial use as Sectra will incorporate it in future product development. Follow-up financing for continued refinement and validation has also been approved. Other results include academic production in the form of publications and contributions to two PhD theses, as well as an in-depth problem description and an intelligence survey of the area.

Planned approach and implementation

The project has had a clear structure where a refined problem description and an intelligence study has been the foundation for the development effort. With this solid base the project has not had to make larger changes in direction. Moreover, the setup focused on evaluation of the prototype by end users (physicians). This both gave valuable feedback and kept the project from getting stuck in non-applicable research results. The cooperation between the partners has been based on individuals on both sides working closely together, in hindsight a necessity to reach the goals.