iDSL: Automated Performance Prediction and Analysis of Medical Imaging Systems

Abstract

iDSL is a language and toolbox for performance prediction of Medical Imaging Systems; It enables system designers to automatically evaluate the performance of their designs, using advanced means of model checking and simulation techniques under the hood, and presents results graphically. In this paper, we present a performance evaluation approach based on iDSL that (i) relies on few measurements; (ii) evaluates many different design alternatives (so-called “designs”); (iii) provides understandable metrics; and (iv) is applicable to real complex systems. Next to that, iDSL supports advanced methods for model calibration as well as ways to aggregate performance results. An extensive case study on interventional X-ray systems shows that iDSL can be used to study the impact of different hardware platforms and concurrency choices on the overall system performance. Model validation conveys that the predicted results closely reflect reality.

This research was supported as part of the Dutch national program COMMIT, and carried out as part of the Allegio project under the responsibility of the ESI group of TNO, with Philips Medical Systems B.V. as the carrying industrial partner.