3D Ultra-Short Echo Time Magnetic Resonance Imaging for Characterizing Cortical Bone Health

Introduction:: Bone is a dynamic composite biomaterial composed of mineral, organic matrix, and water. Each contributes to bone quality and mechanical integrity and can change independently or together due to disease and in response to treatment. However, current conventional bone imaging modalities rely on X-ray radiation to attenuate the bone’s mineral and thus cannot fully characterize all aspects of bone health. Previous pre-clinical magnetic resonance imaging (MRI) studies suggest that novel 3D ultra-short echo-time MRI (UTE MRI) can innocuously quantify bone water, matrix organization and also provide mineral-based outcomes such as cortical porosity and bone mineral density (BMD) based on MRI’s signal intensity [1]. These comprehensive measurements are significant since X-ray-based modalities (e.g. high-resolution peripheral quantitative computed tomography, HR-pQCT) can calculate cortical porosity and BMD but not bone water and matrix organization. In this project, our first aim was to generate a method for multi-modal comparison of the two unique imaging modalities for bone assessment. The second aim was to establish UTE-MRI bone mineral-based measures and determine if correlations exist with HR-pQCT. We hypothesize that UTE-MRI at clinical field strength (3.0 T) will be able to innocuously serve as an alternative or adjuvant for X-ray based imaging for determining cortical porosity and BMD in femoral bone specimens.

Materials and Methods:: Ten fresh cadaveric femoral bone specimens (donor age: 83.5 +/- 10.12 yrs) were harvested and a 5 cm long section of the proximal femoral shaft was sectioned, cleaned of soft-tissue/marrow, and underwent UTE-MRI and HR-pQCT (Fig1.A-B). 3D radial UTE-MRI was acquired at 3.0T (PrismaFit, Siemens) using an 18-channel body coil (Siemens) with the bone fully submerged in a fluoroinert liquid (Fomblin) to maintain hydration and reduce susceptibility artifact using a short (TE1) and long (TE2) echo time and the following parameters: flip angle: 12°, FOV: 160 mm³, 2 averages, 50K radial spokes, readout BW: 250 kHz, 0.45 mm isotropic resolution, TA: 20min 2s. Femora were scanned with the following phantoms: pure water, calcium hydroxyapatite (CaHA) for BMD analysis, and a series of D₂O-doped water for forthcoming bone water assessment. HR-pQCT was acquired with the bone fully wrapped in phosphate buffered saline (PBS) soaked gauze using a Scanco XtremeCT II (0.06 mm isotropic resolution). Multi-modal images were registered (AnalyzeDirect) and a structure-based mask was applied to the HR-pQCT image (MATLAB) creating a 4.5 mm region of interest (ROI) that could be applied across the multi-modal images per specimen. For UTE-MRI, the porosity index [2], an MRI-based porosity measurement derived from the ratio of TE2 and TE1, and a novel signal-intensity-based MRI surrogate measure for BMD [1] was calculated (MATLAB). Cortical porosity and cortical pore volume were calculated from the HR-pQCT images using established methodology (CTAn, Bruker). Pearson correlation coefficients were used to determine relationships between age and bone outcomes (GraphPad Prism).




Results, Conclusions, and Discussions:: Results: For our first aim, we were able to establish a process to analyze multi-modal bone images and extract mineral-based cortical bone measures from each unique modality. Excitingly, both cortical pore volume and percent porosity (both measured via HR-pQCT, our gold standard) significantly and positively correlated with porosity index measured using the novel UTE-MRI approach (Fig1.C-D). While UTE-MRI’s signal-intensity-based surrogate measure for BMD significantly correlated with specimen age (Fig1.E), work is underway to determine if the UTE-MRI BMD correlates with BMD derived from an X-ray-based approach (HR-pQCT). Finally, there was no significant correlation observed between pore volume (using our gold standard approach, HR-pQCT) and donor age in our cohort of ten specimens (Fig1.F).

Conclusions and Discussion: UTE-MRI was able to perform in excellent agreement with the gold standard HR-pQCT for cortical porosity measures but was able to do so without harmful ionizing radiation. The method for analyzing the UTE MRI data was significantly less computationally time consuming due in part to scan size being in megabytes as opposed to the gigabytes of memory required for the HR-pQCT scans. While age significantly correlated to the BMD calculated with the UTE-MRI scans, work is underway to determine if BMD calculated with UTE-MRI can be comparable to the BMD from HR-pQCT scan. More work will need to be done to determine the extent of which the UTE MRI can be used interchangeably with the HR-pQCT. We are addressing complications that inhibited us from obtaining BMD from the HR-pQCT at this time. Beyond this, the UTE-MRI echo times used will also be altered (from 2.8 ms to 3.8, 4.8 and 6.0 ms) to determine if they influence the correlations found. Additionally, a code is being developed to divide the cortical bone into quadrants to determine how outcomes are impacted due to different loading patterns naturally experienced by the bone. Further, work is underway to extract information about bone water content and organic matrix organization from the UTE MRI. The sum of this work will describe a comprehensive set of bone imaging biomarkers using a single innocuous imaging modality to characterize bone health.

Acknowledgements (Optional): :

References (Optional): : [1] Surowiec RK, Ram S, Idiyatullin D, et al. In vivo quantitative imaging biomarkers of bone quality and mineral density using multi-band-SWIFT magnetic resonance imaging. Bone. 2021.

[2] Jones BC, Jia S, Lee H, Feng A, Shetye SS, Batzdorf A, Shapira N, Noël PB, Pleshko N, Rajapakse CS. MRI-derived porosity index is associated with whole-bone stiffness and mineral density in human cadaveric femora. Bone. 2021.