TL;DR
This paper introduces a novel 3D single-shot Chemical Exchange Saturation Transfer (CEST) MRI sequence using the True Fast Imaging with Steady-State Precession (True FISP) readout. This new sequence significantly improves signal-to-noise ratio (SNR) and image quality compared to existing methods, making it a promising tool for faster and more detailed clinical CEST imaging.
What Problem Does This Paper Address?
Chemical Exchange Saturation Transfer (CEST) MRI is a powerful technique for indirectly detecting metabolites and proteins in vivo. To capture the necessary spectral information, CEST imaging typically requires acquiring multiple images at different frequency offsets. This often necessitates fast imaging sequences to keep scan times within clinical feasibility. However, commonly used fast sequences like Fast Low Angle Shot (FLASH) and Echo Planar Imaging (EPI) suffer from limitations such as low SNR or image distortions due to B0 field inhomogeneity. This paper addresses the need for a faster, high-quality 3D CEST imaging method that overcomes these limitations.
Methods
The core innovation is the development of a 3D single-shot CEST sequence utilizing a True Fast Imaging with Steady-State Precession (True FISP) readout. True FISP, also known as balanced steady-state free precession (bSSFP), is a gradient-recalled echo (GRE) variant known for its efficient magnetization refocusing and higher signal. The authors optimized the scanning parameters for this True FISP-based CEST sequence through simulations. They then compared its performance against the conventional FLASH-based CEST sequence, maintaining identical pre-saturation modes, repetition time (TR), echo time (TE), and scan time. The study involved magnetization simulations, phantom experiments (egg white), and in vivo scans of healthy volunteers and brain tumor patients.
Datasets
The study did not introduce new benchmark datasets. Instead, it utilized:
- An egg white phantom as a model for studying APTw and NOE effects.
- Data from ten healthy volunteers.
- Data from two patients with brain tumors (glioblastoma, WHO grade IV).
Key Insights and Ideas
- True FISP's inherent SNR advantage: The balanced nature of True FISP leads to more efficient steady-state magnetization and thus higher signals, especially at short TRs, which is crucial for fast imaging.
- Improved image clarity: The True FISP readout provides sharper images with better delineation of anatomical structures, particularly in the context of CEST contrast, compared to the blurring observed with FLASH.
- Potential for faster, higher-resolution imaging: The enhanced SNR of True FISP opens doors for achieving higher spatial resolution or further accelerating scan times in 3D CEST imaging, which is vital for clinical applications.
- Consistent CEST effects: Despite the improved SNR and image quality, the fundamental CEST effects (like APTw and NOE values) remained comparable between True FISP and FLASH, suggesting that True FISP provides a better readout of these existing phenomena.
Key Results
- Simulation experiments: Showed significantly higher maximum integrated CEST signal with True FISP compared to FLASH.
- Egg white phantom:
- APTw and NOE map SNRs were 68.3% and 57.0% higher with True FISP, respectively.
- APTw and NOE map CNRs increased by 134.4% and 182.8% with True FISP, respectively.
- Healthy volunteers:
- Mean temporal SNR (tSNR) of saturated images at 3.5 ppm was approximately 84% higher with True FISP.
- True FISP displayed finer anatomical structures and clearer tissue contrast.
- Brain tumor patients:
- True FISP revealed a clearer boundary of tumor tissue compared to the more blurred appearance with FLASH.
- APTw (2 uT) images showed tumor hyperintensity and near-zero background signal in normal tissue.
- Scan time: The 3D single-shot CEST imaging with True FISP readout was achieved in approximately 6.1 minutes.
Limitations and Future Work
- SAR: The True FISP sequence, due to its larger flip angle for optimal signal, resulted in higher Specific Absorption Rate (SAR) values (62%) compared to FLASH (6%). The authors suggest using asymmetric RF pulses in future studies to reduce SAR.
- B0 inhomogeneity and susceptibility: True FISP is sensitive to B0 inhomogeneity, which can lead to banding artifacts, particularly in body imaging. Strategies like active shimming, shortest TR, fat-saturation, and specialized True FISP sequences (e.g., CISS) are suggested for mitigation.
- Motion sensitivity: True FISP sequences can be sensitive to head motion, potentially introducing signal instability and noise. Future studies should evaluate and mitigate this motion sensitivity.
- T1 correction: T1 correction was not performed due to the time cost of traditional methods. The authors acknowledge that the absence of T1 correction might affect tumor CEST contrast, potentially leading to overestimation of APTw and NOE signals.
- B1 correction: B1 correction was also not applied, which could impact accuracy, though it's less critical at 3T for head imaging. They recommend using B1 field correction in future clinical scans.
- Four-pool Lorentz fitting: This method might lead to overlapping APT effects with fast exchange effects. More advanced analysis methods like polynomial and Lorentzian line-shape fitting (PLOF) or double saturation power techniques are suggested for more accurate quantification.
- Future objectives: The authors aim to improve resolution and achieve whole-brain CEST imaging with high SNR by applying advanced post-processing algorithms.
Summary updated