syngo.via RT Image Suite

The syngo.via RT Image Suite (SOMARIS/8 VB60) is a software solution designed to support comprehensive radiation therapy treatment preparation through advanced multi-modality image visualization, manipulation, and contouring. It employs sophisticated image processing tools and incorporates a novel deep-learning algorithm for generating synthetic CT images from MRI data, aiming to enhance accuracy and workflow efficiency in radiotherapy planning.

The syngo.via RT Image Suite, software version SOMARIS/8 VB60, is an advanced image analysis software designed for radiation therapy planning. Its primary purpose is to facilitate the viewing, manipulation, 3D and 4D visualization, and comparison of medical images from diverse modalities such as CT, PET, MRI, and Linac CBCT. The software enables clinicians to perform crucial pre-treatment tasks, including the segmentation of tumors and organs-at-risk, manual and software-assisted contouring of volumes of interest, identification of points of interest like the isocenter, image registration, and the creation of simple geometric treatment plans, with final results exportable to treatment planning systems. This updated version introduces several modifications to existing features, enhancing usability and functionality in Reference Point Management, Patient Marking, Routine Contouring (with a new structure-cutting tool), and Structure Set Management. A significant advancement is the replacement of the atlas-based algorithm for generating synthetic CT images from MR acquisitions (for brain and pelvis) with a new deep-learning algorithm. Performance data for this AI-based synthetic CT demonstrates high geometric accuracy, with average body outline deviations less than 1 mm, and superior Hounsfield Unit (HU) accuracy compared to the predicate device, achieving values well within 50 HU for soft tissue and 200 HU for bone tissue. General workflow and parallel image display features have also been improved.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the deep-learning algorithm for synthetic CT generation vary across different MRI sequences and patient demographics beyond brain and pelvis applications?; What specific usability improvements in Reference Point Management and Structure Set Management have demonstrated the most significant impact on clinical workflow efficiency?; Can the new tool for cutting a structure at a user-defined axial level in Routine Contouring be integrated with automated planning workflows, and what is its precision?; What are the computational time and resource requirements for the deep-learning based synthetic CT generation compared to the previous atlas-based method, particularly for brain and pelvis cases?; Are there plans to expand the deep-learning synthetic CT capabilities to other anatomical regions or to incorporate additional MR sequences for improved tissue differentiation and HU accuracy?