Automatic SNR-driven quality control of 2D Phase-Contrast MR imaging using closed-loop feedback between image reconstruction and acquisition

Authors

• A Saurabh
• HA Tarfa
• Y Baumer
• A Dave
• MA Pita
• MA Citron
• LR Ortiz-Whittingham
• S Reynolds
• KA Potharaju
• AS Baez
• K Thompson
• FO Baah
• JF Troendle
• C Ayers
• SJ Neally
• VM Mitchell
• A Wells
• M Marah
• BS Collins
• TM Powell-Wiley

Abstract

Conventional MRI uses predefined parameters with a fixed acquisition time to provide suitable image quality for most patients. Nevertheless, image quality is patient-dependent. Therefore, we propose using a closed-loop feedback framework between data acquisition and the image reconstruction to efficiently achieve consistent diagnostic image quality for quantitative flow measurements with MRI.
The workflow is designed as follows: the “FIRE” framework (Siemens Healthcare) handles the communication between the image acquisition software and the Gadgetron reconstruction software, which rapidly generates images. When the target signal-to-noise ratio (SNR) is achieved, a message is sent from Gadgetron to the acquisition software to stop the data acquisition.
A volunteer was imaged on a 0.55T Free.Max MRI scanner (Siemens Healthineers, Erlangen, Germany). The flow measurement sequence was run with or without closed-loop feedback (every 20s), with a total duration of 4min30s. A target SNR of 120 in the aorta was selected to demonstrate the stopping criterion.
The acquisition stopped when the target SNR was reached after 140s, saving >2 minutes of scan time. The absolute relative difference of quantitative parameters (peak flow and cardiac output) was <5 % between acquisitions with or without feedback.
We have demonstrated a proof-of-concept framework for adaptive MRI and applied it to SNR quality control imaging. It can improve MR value by removing inefficiencies in imaging by reducing scan time or avoiding sequence repetition due to poor image quality. This framework will be extended for diverse applications such as real-time image quality assessment and adaptive sampling for efficient dynamic imaging.

Scientific Focus Area: Biomedical Engineering and Biophysics

This page was last updated on Monday, September 25, 2023