Percentage fat fraction in magnetic resonance imaging: upgrading the osteoporosis-detecting parameter

Rong Chang, Xiaowen Ma, Yonghong Jiang, Dageng Huang, Xiujin Chen, Ming Zhang, Dingjun HaoBMC Medical Imaging 2020 · DOI

TL;DR

This study explores the potential of magnetic resonance spectroscopy (MRS) and m-Dixon-Quant MRI techniques to detect osteoporosis by measuring bone marrow fat fraction (FF%). The findings suggest that FF% is a promising, non-invasive parameter that correlates negatively with bone mineral density (BMD) and is significantly higher in individuals with osteoporosis, offering a potential new avenue for early diagnosis and monitoring.

What Problem Does This Paper Address?

Osteoporosis (OP) is a widespread metabolic bone disorder characterized by reduced bone density and increased fragility, leading to a significant global health burden. Current diagnostic methods, primarily dual-energy X-ray absorptiometry (DXA), have limitations. DXA primarily measures bone mineral density (BMD), which only reflects a portion of bone health and can misclassify individuals, especially older postmenopausal women. Furthermore, DXA has ionizing radiation and limitations in differentiating healed from non-healed fractures. There's a clear need for non-invasive, more comprehensive methods to assess bone quality and predict fracture risk. This paper investigates whether magnetic resonance (MR) techniques, specifically measuring bone marrow fat fraction (FF%), can serve as a more sensitive and specific indicator of osteoporosis compared to traditional BMD measurements.

Methods

The study employed two MR-based techniques to quantify bone marrow fat fraction (FF%):

  • 1-H Magnetic Resonance Spectroscopy (MRS): This technique measures the chemical composition of tissues by detecting resonance frequencies. For this study, it was used to quantify the fat and water peaks in the vertebral body, allowing for the calculation of FF%.
  • m-Dixon-Quant Imaging: This is a more recent MRI technique that acquires multiple echoes at different time points to differentiate fat and water signals. It offers a faster acquisition time and direct quantification of fat fraction.

Both methods were applied to the L3 vertebral body. Bone mineral density (BMD) was measured using quantitative computed tomography (QCT) as a reference standard. The study also collected demographic data (gender, age, height, weight, waist, and hip circumference) to explore their correlation with BMD and FF%. Statistical analyses included Pearson's correlation for relationships between variables, one-way ANOVA for group comparisons, and unpaired t-tests for gender differences.

Datasets

The study enrolled 76 subjects. These subjects were divided into three groups based on their QCT-measured BMD:

  • Normal control group (18 cases)
  • Osteopenia group (30 cases)
  • Osteoporosis (OP) group (28 cases)

The demographic data and MR measurements (FF% from both 1H MRS and m-Dixon-Quant) were collected for each participant. No new benchmark datasets were introduced; the study utilized existing patient data.

Key Insights and Ideas

  • Bone Marrow Fat as a Biomarker: The central idea is that increased fat content in the bone marrow is a physiological change associated with osteoporosis. As bone density decreases, the marrow space may be increasingly filled with fat cells, reflecting a shift from active bone formation to fat accumulation.
  • FF% as a Superior Indicator: The study proposes that FF%, as measured by MR techniques, might be a more sensitive indicator of osteoporosis than BMD alone, potentially detecting changes before significant BMD loss occurs.
  • Non-Invasive Assessment: The use of MRI for this assessment is a significant advantage, offering a non-ionizing radiation method for evaluating bone health, which is crucial for widespread screening and monitoring.
  • Demographic Correlations: The findings highlight that age and gender are important factors influencing both BMD and FF%, suggesting that these demographic parameters should be considered in the context of OP diagnosis and prediction.

Key Results

  • Negative Correlation with BMD: Both 1H MRS and m-Dixon-Quant FF% values showed a significant negative correlation with BMD. This means as BMD decreases, FF% increases.
  • Higher FF% in OP Group: The FF% values, particularly from the m-Dixon-Quant method, were significantly higher in the osteoporosis group compared to the normal control group.
  • Age and FF% Correlation: FF% from m-Dixon scans positively correlated with age, while BMD showed a negative linear relationship with age.
  • Gender Differences: Females exhibited significantly higher FF% values compared to males.
  • Height and BMD Correlation: Height was positively correlated with BMD but not significantly with FF%.
  • Diagnostic Potential: ROC analysis suggested that FF% could diagnose OP with high sensitivity and specificity, potentially up to 100% in distinguishing normal from OP groups.

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Size: The study's sample size was relatively small, particularly for the normal control group in the 1H MRS analysis.
  • Lack of DXA Reference: DXA, the current gold standard for OP diagnosis, was not performed on these patients, limiting direct comparison.
  • QCT Limitations: QCT can have increasing fat errors with age.
  • Limited Prospective Data: The study was cross-sectional. The diagnostic value of FF% needs to be demonstrated in a large prospective cohort.

Future work should focus on:

  • Larger Prospective Studies: Validating these findings in larger, prospective cohorts to confirm the diagnostic and predictive value of FF%.
  • Integration with Other Modalities: Exploring the combined use of MRI techniques with QCT or DXA for a more comprehensive assessment of bone health.
  • Clinical Application: Further research is needed to establish FF% as a routine clinical parameter for OP diagnosis and management.

Summary updated

BibTeX
@article{chang2020percentage,
  author = {Rong Chang and Xiaowen Ma and Yonghong Jiang and Dageng Huang and Xiujin Chen and Ming Zhang and Dingjun Hao},
  title = {{Percentage fat fraction in magnetic resonance imaging: upgrading the osteoporosis-detecting parameter}},
  journal = {BMC Medical Imaging},
  year = {2020},
  doi = {10.1186/s12880-020-00423-0},
}
Figure