Global and peak local specific absorption rate control on parallel transmit systems using k-means SAR compression...

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Biomedical Engineering
• Computational Imaging

Background:

• Parallel transmission (pTx) MRI systems require accurate specific absorption rate (SAR) estimation for safety and performance.
• Existing SAR compression models, like the virtual observation point (VOP) model, have limitations in accuracy and control.

Purpose of the Study:

• To enhance the capability of SAR compression models in pTx MRI systems.
• To develop a more accurate and controlled method for estimating peak local SARs.

Main Methods:

• A k-means clustering algorithm was employed to group voxels with similar SAR behaviors.
• The proposed k-means compression model was compared against the conventional VOP model within a pTx MRI framework.
• Simulations were performed using a numerical human head model and an eight-channel 7T coil array, evaluating RF power, global/peak local SARs, and excitation accuracy.

Main Results:

• The k-means compression model demonstrated a narrower overestimation bound, resulting in more precise local SAR estimation.
• This model achieved the optimal balance between SAR control and excitation accuracy among various pTx pulse design strategies.
• A moderate increase in maximum RF power was found to be beneficial for reducing maximum local SAR deposition.

Conclusions:

• The developed k-means SAR compression model offers significant advantages for pTx MRI, including improved SAR estimation accuracy and adjustable compression ratios.
• This model provides better control over SAR deposition, enhancing the safety and efficacy of pTx MRI applications.
• Findings suggest that optimizing RF power can be a viable strategy for mitigating local SAR levels.