Evaluation of adaptive combination of 30‐channel head receive coil array data in ²³<scp>N</scp>a <scp>MR</scp> imaging

Abstract

PURPOSE: The aim was to optimally combine multichannel coil array data in sodium ((23) Na) MRI. METHODS: (23) Na MRI was conducted on a 3 Tesla MR system using a 30-channel head receive coil array. The parameters used for the adaptive combination (ADC) reconstruction of the low signal-to-noise ratio (SNR) dataset have been optimized by finding the maximum mean SNR. A pseudo multiple-replica approach has been used to obtain SNR maps of the combined images. To prove reproducibility of the combination algorithm, the procedure was repeated for several measurements. RESULTS: For low SNR data, sum-of-squares (SOS) reconstruction leads to high background noise and a signal bias in the imaged object. The ADC reconstruction clearly reduces noise in the image and leads to an increase of the mean SNR in the range of 8% to 50%, compared to weighted SOS depending on the absolute SNR of the image. The evaluation of the effects of different noise scans showed that a small number of projections can be used to estimate noise statistics of the coil array without substantially decreasing the resulting SNR. CONCLUSION: (23) Na MRI can be markedly improved by using a 30-channel receive array and ADC reconstruction. The ADC reconstruction showed robust results for all measurements without the need for sensitivity maps.