VEA Align

VEA Align is a cloud-based software intended to assist healthcare professionals in orthopedic global alignment assessment using 2D X-ray images from EOS systems. It employs a machine learning algorithm for initial anatomic landmark placement, which users validate and adjust to derive precise clinical measurements for analyzing spinal and lower limb conditions.

VEA Align is a cloud-based software designed for orthopedic applications in both pediatric (aged 7 years and older) and adult populations. Its primary purpose is to assist healthcare professionals in assessing patient-specific global alignment and aid in the analysis of spinal deformities, degenerative diseases, and lower limb alignment disorders through precise angle and length measurements. The system utilizes biplanar 2D X-ray images exclusively acquired from EOS imaging's EOS and EOSedge systems. Upon image input, a machine learning-based algorithm proposes an initial placement of anatomic landmarks on the frontal and lateral images, which users can then interactively adjust and must validate. The core science of VEA Align revolves around the computation of clinical parameters derived from these user-validated anatomic landmarks, which are recalculated in real-time as adjustments are made; these parameters may also be compared to published normative values for additional assessment. The software operates on a cloud infrastructure, supported by VEA Portal for user access control and data management. Nonclinical performance testing confirmed the device's design, usability, and the reproducibility and accuracy of its clinical parameter outputs. Notably, the machine learning algorithm's performance was validated using a dataset of 555 patients, demonstrating that its landmark identification met acceptance criteria when compared to the predicate sterEOS Workstation.

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. What specific types of anatomic landmarks does the machine learning algorithm identify, and what is its reported accuracy for each landmark compared to manual placement by experienced clinicians, beyond the comparison to the sterEOS workstation?; How does VEA Align's cloud-based architecture ensure data security and HIPAA compliance, particularly when transmitting and storing patient 2D X-ray images and derived parameters?; Could the system be expanded to incorporate or correlate findings with other imaging modalities (e.g., CT, MRI) or clinical data for a more comprehensive patient assessment?; What are the "published normative values" referenced for comparison, and how are they integrated and updated within the software to ensure relevance across diverse patient populations (e.g., different ethnicities, age groups beyond the 7-year minimum)?; What was the range and diversity of pathologies (spinal deformities, degenerative diseases, lower limb disorders) included in the 555-patient dataset used for validating the machine learning algorithm's performance?