Veuron-Brain-pAb3

Veuron-Brain-pAb3 is medical imaging software intended to support dementia assessment through quantitative analysis of amyloid protein concentration using PET and MRI brain scans. The system employs image registration, fusion, and a Convolutional Neural Network for segmentation, calculating the Standardized Uptake Value Ratio (SUVR) with selectable beta-amyloid tracer options to aid clinical diagnosis.

Veuron-Brain-pAb3 is a standalone medical imaging software designed to assist clinicians in the assessment and quantification of brain pathologies, specifically from PET amyloid scans relevant to dementia diagnosis. The software processes multimodal medical images, including MRI and PET scans, by performing registration, fusion, display, and analysis. Its core function is the quantitative analysis of PET amyloid data through the automatic calculation of the Standardized Uptake Value Ratio (SUVR), providing a reference to support medical professionals in their diagnostic interpretations. The system is deployed in medical imaging workplaces and utilizes workflows tailored for specific radio-tracer and disease combinations, offering user-selectable options for processing various PET amyloid images. The technological foundation of Veuron-Brain-pAb3 centers on loading MR and PET images, overlaying them, and calculating SUVR for brain regions to visualize amyloid protein concentration by comparing target regions to reference regions. Image segmentation and volume calculation are achieved using a Convolutional Neural Network (CNN) model, which was trained on 3D brain MR images. Key functionalities include importing DICOM and NIfTI data, automatic post-processing, and exporting results as PNG images or CSV/PNG reports. Enhancements over its predicate device include the addition of a patient worklist and a significant tracer option feature, allowing selection from specific beta-amyloid tracers (18F-Flutemetamol, 18F-Florbetapir, 18F-Florbetaben) to refine analyses. Non-clinical testing confirmed that these new functions perform as intended.

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 selection of different beta-amyloid tracers (18F-Flutemetamol, 18F-Florbetapir, 18F-Florbetaben) impact the SUVR calculation methodology and the subsequent clinical interpretation for amyloid quantification?; What were the performance metrics (e.g., accuracy, sensitivity, specificity) of the Convolutional Neural Network (CNN) model used for brain segmentation, and how was it validated across diverse patient populations, particularly given it was trained on images from a single domestic institution?; Can the software integrate seamlessly with various existing PACS/RIS systems, and what are the specific technical requirements for its deployment and maintenance in a typical hospital's medical imaging workplace?; How does the Veuron-Brain-pAb3 define and delineate 'target regions of interest' and 'reference regions' for SUVR calculation, and does it offer clinicians the flexibility to manually adjust or define these regions if needed?; What specific non-clinical performance tests were conducted to validate the accuracy of SUVR calculations and the reliability of the new worklist and tracer option features, especially concerning their impact on diagnostic support compared to the predicate device?