Optimizing Transmit Field Inhomogeneity of Parallel RF Transmit Design in 7T MRI using Deep Learning

2024-08-21

arXiv.org

AI Summary

B₁⁺ field inhomogeneity at 7T MRI causes flip-angle deviations and image artifacts, limiting clinical utility. This paper proposes a two-stage deep learning framework for rapid and accurate optimization of multi-channel RF parallel excitation. We introduce an end-to-end residual learning architecture that eliminates the need for precomputed reference shim weights, thereby removing dependence on patient-specific in-scan measurements and enhancing both generalizability and real-time feasibility. Leveraging Adam-initialized ResNet, the model directly learns the mapping from multi-channel B₁⁺ field maps to optimal shim weights. Compared with conventional matrix least-squares (MLS) optimization, our method accelerates shim computation by over one order of magnitude while significantly improving B₁⁺ uniformity. Consequently, image signal-to-noise ratio (SNR) and contrast are simultaneously enhanced. The proposed approach establishes a robust, efficient, and real-time RF calibration paradigm for 7T MRI.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) provides a higher signal-to-noise ratio and, thereby, higher spatial resolution. However, UHF MRI introduces challenges such as transmit radiofrequency (RF) field (B1+) inhomogeneities, leading to uneven flip angles and image intensity anomalies. These issues can significantly degrade imaging quality and its medical applications. This study addresses B1+ field homogeneity through a novel deep learning-based strategy. Traditional methods like Magnitude Least Squares (MLS) optimization have been effective but are time-consuming and dependent on the patient's presence. Recent machine learning approaches, such as RF Shim Prediction by Iteratively Projected Ridge Regression and deep learning frameworks, have shown promise but face limitations like extensive training times and oversimplified architectures. We propose a two-step deep learning strategy. First, we obtain the desired reference RF shimming weights from multi-channel B1+ fields using random-initialized Adaptive Moment Estimation. Then, we employ Residual Networks (ResNets) to train a model that maps B1+ fields to target RF shimming outputs. Our approach does not rely on pre-calculated reference optimizations for the testing process and efficiently learns residual functions. Comparative studies with traditional MLS optimization demonstrate our method's advantages in terms of speed and accuracy. The proposed strategy achieves a faster and more efficient RF shimming design, significantly improving imaging quality at UHF. This advancement holds potential for broader applications in medical imaging and diagnostics.

Problem

• Addresses B1+ field inhomogeneity in 7T MRI
• Proposes deep learning for RF shimming optimization
• Improves imaging quality and efficiency in UHF MRI

Innovation

• Deep learning optimizes RF shimming
• Residual Networks map B1+ fields
• Adaptive Moment Estimation initializes weights