SIGNA 7.0T

The SIGNA 7.0T is a high-field magnetic resonance imaging system designed for diagnostic imaging of the head and extremities, now incorporating the AIR Recon DL software feature to enhance image quality. This feature utilizes deep learning-based reconstruction to improve signal-to-noise ratio and image sharpness, as validated by non-clinical and clinical studies showing superior image characteristics and radiologist preference.

The SIGNA 7.0T is a whole-body magnetic resonance scanner operating at a 7.0 Tesla field strength, designed to produce high-resolution images for diagnostic purposes. It supports various imaging techniques including axial, sagittal, coronal, and oblique views, as well as spectroscopic images, parametric maps, and dynamic images of the head and extremities. The system utilizes a superconducting magnet and an ultra-high performance gradient coil within a 60 cm patient bore. The images generated reflect the spatial distribution or molecular environment of nuclei exhibiting magnetic resonance, providing information that, when interpreted by a trained physician, can assist in diagnosis for patients weighing over 20 kg (44 lb). This particular iteration of the SIGNA 7.0T introduces the AIR Recon DL software feature and includes upgrades for installed base magnet systems. AIR Recon DL is a deep learning-based image reconstruction technology intended to improve image quality by increasing signal-to-noise ratio (SNR) and image sharpness while reducing image noise. Non-clinical performance testing demonstrated these improvements, showing that AIR Recon DL can maintain image SNR and sharpness even with reduced scan times, without adversely affecting low contrast detectability. Clinical evaluations, including a reader study with radiologists, confirmed that AIR Recon DL provides images with better SNR and equivalent or superior sharpness, leading to a uniform preference for these images in clinical assessment.

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 AIR Recon DL algorithm differentiate between true anatomical signal and noise, particularly for subtle pathologies in high-resolution 7.0T imaging?; What is the quantifiable reduction in scan time achievable with AIR Recon DL on the SIGNA 7.0T while maintaining or improving diagnostic image quality compared to conventional reconstruction methods?; Are there specific contraindications or patient populations (e.g., those with certain implants compatible at 7T but perhaps interacting differently with the DL reconstruction) for whom the AIR Recon DL feature might be less suitable or require specific parameter adjustments?; How does the performance of AIR Recon DL on the SIGNA 7.0T compare to its previously cleared implementation on 3T systems, considering the differences in field strength, SNR characteristics, and potential artifacts?; What training data was used for the AIR Recon DL model specific to 7.0T imaging, and how does it account for the unique image contrasts and artifacts encountered at ultra-high field strengths?