Towards Penile Small Vessel Imaging with Ferumoxytol-Enhanced MRI 📝

Author: Darren Fang, Amar Kishan, Justin McWilliams, Dan Ruan, Xiaodong Zhong 👨‍🔬

Affiliation: Department of Radiation Oncology, University of California, Los Angeles, Department of Radiology, University of California, Los Angeles, Department of Radiological Sciences, University of California, Los Angeles 🌍

Abstract:

Purpose: Prostate radiotherapy can malform penile vasculature, contributing to erectile dysfunction and compromising quality of life. To detect, quantify, and preferably avoid such occurrences, this project aims to (1) optimize penile MR angiography protocols and (2) develop physical and anthropomorphic phantoms to characterize imaging performance.

Methods: To image penile arteries which reach the MRI resolution limit, we investigate T1 MP-RAGE and T2 TSE sequences with intravascular contrast from ferumoxytol at 1-4 mg/kg. Resolution and vessel conspicuity are optimized using custom-built physics and patient-derived anthropomorphic phantoms. The physics phantom consists of tubes of diameters ranging from 0.5 to 6 mm mounted with varying line-pair spacings on a 3D-printed frame. Various ferumoxytol concentrations were investigated. Additional anthropomorphic models were developed to reflect anatomical variation and vessel tortuosity. Under IRB approval, penile vasculature skeletons were segmented from CT angiographies, 3D-printed, and fitted with ferumoxytol-loaded tubing. Both phantom sets were scanned in a soft tissue-equivalent NiCl2/MnCl2 solution (T1~1400 ms) with MP-RAGE and TSE at 1.5 T. Scans are approximately 10 minutes with sub-millimeter in-plane resolution. TSE gives higher contrast-to-noise ratio in less time. Ferumoxytol aids vessel detection with T1 hyperintensity and T2 hypointensity. Resolution and vessel conspicuity were visually assessed and quantified using intensity full width at half maximum (FWHM).
Results: Tubes ≥ 1 mm are accurately visualized with 4 mg/kg ferumoxytol. FWHM approximates millimeter-scale diameters with errors of 17% and 7% from T1 and T2 images, respectively. 0.5 mm tubes approach the MRI resolution limit and are poorly resolved, necessitating super-resolution. 1 mm tube line pairs are distinguishable at distances above 1 mm on T1 and 3 mm on T2. Tortuous trajectories on anthropomorphic phantoms can be recovered from both T1 and T2 scans.

Conclusion: Our investigation demonstrates the promise of meticulously optimized ferumoxytol-enhanced MRA to visualize fine penile vessels supplying erectile function.

Back to List