1350520 - CNNT denoising for cine imaging at 0.55T with higher acceleration rates

Low magnetic fields of 0.35T and 0.55T have demonstrated promise for clinical CMR [1, 2]. Low field systems may potentially increase the accessibility of cardiac MRI by reducing the hardware and operation cost. Low field MRI offers inherently lower SNR, and commercial low field systems operate with reduced gradient performance and fewer receiver channels. These hardware limitations restrict rapid imaging capabilities by enforcing longer TRs and lower accelerations rates (typically R=1 or 2).

Deep learning offers the capability to denoise images, but many denoising approaches are naïve to the non-uniformity of noise amplification based on coil geometry, g-factor, and underlying SNR. Moreover, most algorithms require retraining with large datasets for each field strength.  Here, we present the application of a g-factor-savvy denoising for cine imaging at 0.55T to increase useable acceleration rate.



Methods:

Our model utilizes a new network architecture combining convolutional neural networks (CNNs) and the more recent transformer model. The novel CNN transformer denoising model (CNNT) [3] used both complex 2D+time images in SNR-units [4] and g-factor maps as inputs. The model was trained on 3T cine data, and applied to 0.55T cine without retraining. The denoising model was deployed inline on the scanner using Gadgetron InlineAI [5].

Images were acquired on a commercial 0.55T MRI system (MAGNETOM Free.Max, Siemens Healthcare, Erlangen, Germany). Breath-held retrospectively-gated bSSFP cine images were acquired using acceleration rates R=2, 3, and 4 (parameters in Table 1), with ECG gating by an external physiological monitoring system (Tesla, Surgical Tools Inc, Bedford VA). Imaging was performed on 9 healthy volunteers and one patient with known wall-motion abnormality.

SNR gain between the original and denoised images was measured in the septum in a midventricular short axis slice, by the pseudo-replica method [4].



Results:

Figures 1 and 2 provide example short and long axis images from acceleration rates R=2, 3, and 4. Rate R=2 acquired only 23 phases per heartbeat and myocardial blurring was evident following the interpolation to 30 frames. The CNNT denoising enabled reliable use of rate R=3 for cine imaging in all subjects, whereas R=4 may start to reach the limit of the model where signal can be distinguished from noise. The wall-motion abnormality in one patient was visible following CNNT denoising. The percent mean myocardial SNR gain across 9 healthy volunteers was 97±31% (R=2), 122±22% (R=3), and 476±130% (R=4).



Conclusion:

CNNT-based denoising demonstrated improved image SNR for cine at 0.55T. Importantly, this CNNT method uses the SNR-scaled image reconstruction and g-factor maps for improved generalization, and the CNNT model did not require retraining for application at 0.55T. Improved acceleration rate can be used to increase acquired temporal resolution or shorten breath-holds, and may reduce the hardware requirements for rapid cardiac imaging at low field.