Go-Smart: Web-based Computational Modeling of Minimally Invasive Cancer Treatments

TL;DR

Go-Smart is a web-based platform that enables personalized modeling and simulation of minimally invasive cancer treatments, aiding clinicians in planning and evaluating procedures using patient-specific data and adaptable simulation tools.

Contribution

It introduces a scalable, extensible web environment for creating patient-specific 3D models and simulating treatment outcomes, integrating multiple imaging modalities and device models.

Findings

Effective prediction of treatment outcomes using simulations

High extensibility for incorporating new devices and models

Validation through comparison with follow-up scan data

Abstract

The web-based Go-Smart environment is a scalable system that allows the prediction of minimally invasive cancer treatment. Interventional radiologists create a patient-specific 3D model by semi-automatic segmentation and registration of pre-interventional CT (Computed Tomography) and/or MRI (Magnetic Resonance Imaging) images in a 2D/3D browser environment. This model is used to compare patient-specific treatment plans and device performance via built-in simulation tools. Go-Smart includes evaluation techniques for comparing simulated treatment with real ablation lesions segmented from follow-up scans. The framework is highly extensible, allowing manufacturers and researchers to incorporate new ablation devices, mathematical models and physical parameters.