Symposium



Novel RF Coil Design to Incorporate Parallel Imaging into 13C MRSI for Human Subjects at 3T


Presenting Author Senior Author
Name: Natalie Korn Name: Susan Noworolski
Email: natalie.korn@ucsf.edu Email: susan.noworolski@ucsf.edu
Presenting Author’s RIG/SRG: Prostate Cancer Abdominal Imaging  
Presenting Author's Lab Location: China Basin   

Abstract Information
Imaging Modality: MR
Disease Application: Prostate Cancer
Complete author list: Natalie Korn, John Kurhanewicz, Susan Noworolski
Abstract highlights: Hyperpolarized 13C can show real-time metabolism in prostate cancer, but cannot be applied in the clinic due to hardware limitations. In this work, we design and simulate an improved endorectal coil to allow parallel imaging at the carbon frequency without sacrificing SNR over the prostate gland.
 
Introduction
Prostate cancer is the second most common cancer in American men. However, prostate cancer can be slow-growing, and determining metabolic behavior has emerged as a prominent method of measuring growth potential. 13C magnetic resonance spectroscopic imaging (MRSI) using hyperpolarized carbon-labeled pyruvate probes a specific chemical pathway associated with cancerous energy usage—anaerobic metabolism in the presence of oxygen known as the Warburg Effect. While this is shown to correlate with growth potential in mice, the nonrenewable signal decays too quickly for sufficient spatial localization in human subjects. To increase imaging speed, researchers employ parallel imaging, which requires multiple receiving elements. In prostate imaging, an endorectal coil (ERC) is used to amplify signal-to-noise ratio (SNR) over the prostate, but no ERC has been developed with multiple receive elements to allow parallel imaging at the carbon frequency. In this work, we determine an optimal ERC design with multiple receiving elements tuned to 13C to employ parallel imaging, enabling diagnostic quality 13C MRSI in the clinic at 3T. Because SNR decreases in parallel imaging, ERC geometry will be tailored to maximize SNR over the prostate.
 
Methods
The functional constraint of this work is to incorporate parallel imaging without sacrificing the signal-to-noise ratio (SNR) over the prostate gland. To compare coil element geometries and sensitivity profiles, custom in-house software modeled the magnetic field variation for six potential geometries. Because signal reception depth is dependent on coil size, and signal is only necessary over the prostate gland, 847 recent exams from the UCSF Prostate Imaging Research Group were interrogated for distribution in prostate size. As an ethical constraint, coil design cannot increase the size of the ERC.
 
Results
To cover the gland in 95% of 847 recent exams, signal reception must reach 4.71 cm in A/P and 6.69 cm in R/L. Geometries incorporating a receive array in the R/L direction increased unnecessary signal reception to the right and left of the prostate, while receive arrays in the S/I direction did not. Using a larger element at the base of the prostate with a smaller element at the apex better matched average prostate gland shape, which tapers towards the apex. Using four or more receive elements decreased the reception depth to less than 4.71 cm, resulting in a reception profile that would not cover the prostate. The optimal performance in simulation was a receive array in S/I using three elements, with element width in R/L decreasing towards the apex analogous to prostate shape. This geometry’s signal reception profile satisfied the 95th percentile of gland reception in both A/P and R/L, as shown in Figure 1.
 
Conclusions
Using a novel endorectal coil geometry with a three-element array in S/I with element size decreasing towards the apex can provide parallel imaging capabilities to human MRI exams at 3T without sacrificing SNR over the prostate, translating 13C MRSI of the prostate to the clinical setting.