Radiological Sciences, University of California Los Angeles, Los Angeles, CA, USA.
Urology, University of California Los Angeles, Los Angeles, CA, USA.
MAGMA. 2022 Aug;35(4):667-682. doi: 10.1007/s10334-022-01029-z. Epub 2022 Jul 23.
This study aimed at developing dictionary learning (DL) based compressed sensing (CS) reconstruction for randomly undersampled five-dimensional (5D) MR Spectroscopic Imaging (3D spatial + 2D spectral) data acquired in prostate cancer patients and healthy controls, and test its feasibility at 8x and 12x undersampling factors.
Prospectively undersampled 5D echo-planar J-resolved spectroscopic imaging (EP-JRESI) data were acquired in nine prostate cancer (PCa) patients and three healthy males. The 5D EP-JRESI data were reconstructed using DL and compared with gradient sparsity-based Total Variation (TV) and Perona-Malik (PM) methods. A hybrid reconstruction technique, Dictionary Learning-Total Variation (DLTV), was also designed to further improve the quality of reconstructed spectra.
The CS reconstruction of prospectively undersampled (8x and 12x) 5D EP-JRESI data acquired in prostate cancer and healthy subjects were performed using DL, DLTV, TV and PM. It is evident that the hybrid DLTV method can unambiguously resolve 2D J-resolved peaks including myo-inositol, citrate, creatine, spermine and choline.
Improved reconstruction of the accelerated 5D EP-JRESI data was observed using the hybrid DLTV. Accelerated acquisition of in vivo 5D data with as low as 8.33% samples (12x) corresponds to a total scan time of 14 min as opposed to a fully sampled scan that needs a total duration of 2.4 h (TR = 1.2 s, 32 [Formula: see text]×16 [Formula: see text]×8 [Formula: see text], 512 [Formula: see text] and 64 [Formula: see text]).
本研究旨在开发基于字典学习（DL）的压缩感知（CS）重建方法，用于重建前列腺癌患者和健康对照者随机欠采样的 5 维（3D 空间+2D 谱）磁共振波谱成像（MRSI）数据，并在 8x 和 12x 欠采样因子下测试其可行性。
前瞻性地采集了 9 例前列腺癌（PCa）患者和 3 例健康男性的 5 维 EPI-JRESI 数据。使用 DL 对 5D EP-JRESI 数据进行重建，并与基于梯度稀疏的全变差（TV）和 Perona-Malik（PM）方法进行比较。还设计了一种混合重建技术，即字典学习-全变差（DLTV），以进一步提高重建谱的质量。
使用 DL、DLTV、TV 和 PM 对前列腺癌和健康受试者的前瞻性欠采样（8x 和 12x）5D EP-JRESI 数据进行 CS 重建。显然，混合 DLTV 方法可以明确分辨包括肌醇、柠檬酸盐、肌酸、精胺和胆碱在内的 2D J 分辨峰。
使用混合 DLTV 可明显改善加速 5D EP-JRESI 数据的重建。以低至 8.33%的样本（12x）加速采集体内 5D 数据，总扫描时间为 14 分钟，而完全采样扫描需要 2.4 小时（TR=1.2s，32[公式：见文本]×16[公式：见文本]×8[公式：见文本]，512[公式：见文本]和 64[公式：见文本]）。
