Assessment of Bone Microarchitecture in Fresh Cadaveric Human Femurs: What Could Be the Clinical Relevance of Ultra-High Field MRI.
Enrico SoldatiMartine PithiouxDaphne GuenounDavid BendahanJerome VicentePublished in: Diagnostics (Basel, Switzerland) (2022)
MRI could be applied for bone microarchitecture assessment; however, this technique is still suffering from low resolution compared to the trabecular dimension. A clear comparative analysis between MRI and X-ray microcomputed tomography (μCT) regarding microarchitecture metrics is still lacking. In this study, we performed a comparative analysis between μCT and 7T MRI with the aim of assessing the image resolution effect on the accuracy of microarchitecture metrics. We also addressed the issue of air bubble artifacts in cadaveric bones. Three fresh cadaveric femur heads were scanned using 7T MRI and µCT at high resolution (0.051 mm). Samples were submitted to a vacuum procedure combined with vibration to reduce the volume of air bubbles. Trabecular interconnectivity, a new metric, and conventional histomorphometric parameters were quantified using MR images and compared to those derived from µCT at full resolution and downsized resolutions (0.102 and 0.153 mm). Correlations between bone morphology and mineral density (BMD) were evaluated. Air bubbles were reduced by 99.8% in 30 min, leaving partial volume effects as the only source of bias. Morphological parameters quantified with 7T MRI were not statistically different ( p > 0.01) to those computed from μCT images, with error up to 8% for both bone volume fraction and trabecular spacing. No linear correlation was found between BMD and all morphological parameters except trabecular interconnectivity (R 2 = 0.69 for 7T MRI-BMD). These results strongly suggest that 7T MRI could be of interest for in vivo bone microarchitecture assessment, providing additional information about bone health and quality.
Keyphrases
- contrast enhanced
- bone mineral density
- magnetic resonance imaging
- diffusion weighted imaging
- computed tomography
- magnetic resonance
- dual energy
- high resolution
- postmenopausal women
- image quality
- body composition
- healthcare
- bone loss
- deep learning
- soft tissue
- public health
- positron emission tomography
- single molecule
- endothelial cells
- mental health
- convolutional neural network
- bone regeneration
- optical coherence tomography
- risk assessment
- human health
- machine learning
- finite element