SIGNA Premier

The SIGNA Premier is a 3.0T whole-body magnetic resonance diagnostic device designed to produce high-resolution images for a wide range of anatomical regions and applications. This iteration introduces the AIRx software, a deep learning-based feature that automates and standardizes pre-scan steps for brain MRI examinations to improve workflow productivity and consistency while maintaining established imaging performance.

The SIGNA Premier is a whole-body magnetic resonance (MR) scanner designed for diagnostic imaging, supporting high resolution, high signal-to-noise ratio, and short scan times, while aiming for improved patient comfort and workflow. It utilizes a 3.0T superconducting magnet with a 70cm bore and can produce axial, sagittal, coronal, oblique, and double oblique images, spectroscopic images, parametric maps, and dynamic images of various anatomical structures and functions across the entire body, including head, neck, spine, heart, abdomen, and musculoskeletal regions; contrast agents may be used as needed. The images generated reflect the spatial distribution or molecular environment of nuclei exhibiting magnetic resonance, which, when interpreted by a trained physician, assist in diagnosis. A key modification to this iteration of the SIGNA Premier system is the introduction of the AIRx software feature, which was developed using deep learning algorithms. AIRx is designed to automate and standardize a number of connected steps specifically for MRI examinations of the brain, functioning as a pre-scan algorithm. The purpose of AIRx is to enhance productivity and consistency in brain imaging workflows. Non-clinical testing, including risk analysis and various performance verifications, confirmed compliance with relevant standards like IEC 62304. Clinical validation involved internal scans that confirmed the productivity and consistency benefits of AIRx, with the MR system maintaining the same imaging performance as its predicate device.

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 AIRx deep learning algorithm specifically identify and automate the "connected steps" for brain MRI examinations, and what is the range of variability it can accommodate in patient anatomy or positioning?; What were the specific quantitative metrics used to validate the "productivity and consistency benefits" of the AIRx feature in the internal scans, and how do these compare to workflows without AIRx?; Beyond brain imaging, are there plans to extend the AIRx deep learning technology to automate and standardize MRI examinations for other anatomical regions supported by the SIGNA Premier?; What is the impact of the AIRx pre-scan algorithm on the total examination time for typical brain MRI protocols, and how does it affect the technologist's interaction with the scanner?; Given that AIRx is a pre-scan algorithm, how does it interface with the subsequent pulse sequences and reconstruction algorithms to ensure optimal image quality is maintained across diverse brain pathologies?