syngo.via RT Image Suite

The syngo.via RT Image Suite is a medical image analysis software engineered to improve radiation therapy planning by offering extensive tools for multi-modal image visualization, segmentation, and contouring, significantly featuring AI-powered autocontouring. It employs advanced digital image processing and deep learning algorithms, validated against diverse international clinical data, to enhance the precision and efficiency of preparing treatments like Brachytherapy, Particle Therapy, and External Beam Radiation Therapy.

The syngo.via RT Image Suite, software version SOMARIS/8 VB80, is an advanced image analysis software designed for radiation therapy planning. Its primary purpose is to assist medical professionals by providing comprehensive tools for viewing, manipulating, and visualizing 3D and 4D medical images from diverse modalities including CT, PET, MRI, and Linac CBCT. The system's methodology involves facilitating the critical pre-dosimetric planning steps of segmenting tumors and organs-at-risk through digital image processing and visualization techniques. These techniques support manual and software-assisted contouring, identification of points of interest, multi-modality image registration, and the creation of simple geometric treatment plans, with the final data, such as contours and plans, being exportable to dedicated treatment planning systems for subsequent radiotherapy procedures. This updated version, SOMARIS/8 VB80, introduces specific enhancements, notably to its Advanced Contouring feature and a revised user interface. The science behind the Advanced Contouring relies on a deep learning-based algorithm for automated segmentation, which, while the core algorithm remains unchanged, has now been trained and validated using an expanded and diverse international clinical dataset encompassing varied demographics, imaging parameters, and both native and contrasted CT images to support additional anatomical structures. Non-clinical performance testing results for this AI-driven autocontouring, evaluated on 413 independent subjects by comparing automated segmentations against manual ground-truth contours using the Dice coefficient, demonstrated that the feature met all predefined acceptance criteria for geometric overlap. This confirms the algorithm's safety and effectiveness for its intended clinical application in accurately delineating anatomical structures.

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 extended Advanced Contouring feature (SOMARIS/8 VB80) for the newly added anatomical structures compare quantitatively (e.g., Dice coefficient, Hausdorff distance) to the previously supported structures and to manual contouring by expert clinicians across diverse patient datasets?; What specific "minor modifications" were made to the user interface in SOMARIS/8 VB80, and how do these changes measurably impact workflow efficiency or user error rates in typical radiation therapy planning scenarios?; Given the diverse international data used for training and validation of Advanced Contouring, are there any observable performance variations or limitations of the AI algorithm when applied to patient populations, imaging equipment, or acquisition protocols not extensively represented in the training/validation sets?; What is the typical processing time for the AI-based autocontouring feature for different anatomical regions (e.g., Head & Neck, Thorax & Abdomen, Pelvis) using recommended hardware, and how does this compare to the average time required for manual contouring of the same structures?; How does the accuracy and clinical utility of the generated synthetic CT (sCT) from MR acquisitions compare to standard CT scans for the purpose of radiation therapy dose calculation and plan evaluation, particularly in anatomical sites prone to MR distortions or intensity variations?