Open End MICT Simulation Environment

Workflow Overview

The GO-SMART environment is an open-ended, WEB-based software framework for minimally-invasive cancer treatment (MICT). Within this environment, workflows for planning, simulating and validating MICTs are implemented.

Pre-interventional simulation of MICTs for a given patient comprises the following computational workflow:
  • Image analysis of the patient image data followed by the generation of the patient’s 3D model. This comprises segmentation of patient scans to generate surface meshes for an organ, vessel tree and tumour and to register all the segmented structures into the 3D model of the patient
  • Define simulation MICT set-up such as position of MICT probe and ablation protocol
  • Define simulation parameters and run numerical simulation of ablation lesion

Fig 1: GO-SMART simulation menu as an example for the realization of MICT workflow elements within the GO-SMART environement: The user selects the modality (1), generator (2), protocol (3) and one or more needles (4). Needles can be placed by defining entry (E) and target (T) point and then appear in the viewer (6). After needle placement is finished, selected numerical parameters can be defined (5) and the simulation can be executed.
Post-interventional evaluation of MICTs and validation of numerical models includes the following steps:
  • Registration of the actual applicator positions from inter-operational data and re-running of the numerical model with the actual data.
  • Segmentation of the lesion and validation of the actual lesion agains (i) planning simulation predictions and (ii) results of the numerical model run with actual applicator positions.
  • Lesion-tracking over multiple follow-up scans.


The Go-Smart environment was implemented following a service-oriented approach. The components of the architecture are shown below and comprise:

  • Services to realize image visualization, image processing and simulation.
  • Webserver to handle request and store the data
  • Clients for common browsers or for 3D-Visualiation

The core components of the architecture are implemented as services which run autonomous and can be reused in other application scenarios. To the core components belong the image processing service which performs tasks such as segmentations of anatomical structures, the image visualisation service which generates the images for the frontend, and the simulation service which executes the simulation. It follows a more detailed description of the services.

The distributed architecture allows scaling up the environment by adding additional cloud instances. This is, because each service only communicates with the SignalR hub and the database, and does not store any data locally.