Application of magnetic resonance image compilation (MAGiC) in the diagnosis of middle-aged and elderly women with osteoporosis

Yiming Chen, Xiuting Mei, Xuqian Liang, Yi Cao, Cong Peng, Yang Fu, Yulong Zhang, Cuifang Liu, Yang LiuBMC Medical Imaging 2023 · DOI

TL;DR

This study explores the use of Magnetic Resonance Image Compilation (MAGIC), a novel MRI technique, for diagnosing osteoporosis (OP) in middle-aged and elderly women. By quantifying T1 and T2 relaxation times, MAGIC shows high diagnostic efficiency, potentially offering a radiation-free alternative to traditional methods like DXA.

What Problem Does This Paper Address?

Osteoporosis (OP) is a significant health concern, particularly for middle-aged and elderly women, characterized by decreased bone mineral density (BMD) and increased fracture risk. Current diagnostic methods, such as dual X-ray absorptiometry (DXA) and quantitative computed tomography (QCT), involve radiation exposure. While magnetic resonance imaging (MRI) offers a radiation-free approach with excellent soft-tissue contrast, conventional MRI signals can be influenced by various factors, making direct interpretation of bone marrow composition challenging. Existing MRI techniques for OP evaluation have limitations, including small study populations and a lack of direct correlation with BMD. This paper addresses the need for a safe, effective, and quantitative method to diagnose OP early, focusing on the potential of the MAGIC sequence to provide reliable T1 and T2 relaxation time measurements.

Methods

The researchers employed the Magnetic Resonance Image Compilation (MAGIC) sequence, a quantitative MRI technique that utilizes a multiple-delay multiple-echo (MDME) approach. This method allows for the simultaneous acquisition of quantitative maps for T1 (longitudinal relaxation time) and T2 (transverse relaxation time), along with proton density (PD). These parameters are considered intrinsic properties of tissues and are less susceptible to interference compared to standard MRI signal intensity.

The study involved 110 female patients who underwent both lumbar MAGIC MRI and DXA examinations. The patients were divided into two groups based on their DXA results: an osteoporotic (OP) group and a non-osteoporotic (non-OP) group. The researchers then analyzed the variation trends of T1, T2, and BMD with age. They established a clinical mathematical model to examine the correlation between T1, T2, and BMD. Statistical analyses included t-tests, Pearson correlation coefficients, and Receiver Operating Characteristic (ROC) curve analysis to assess diagnostic accuracy. A Ridge regression model was used to fit a function for BMD based on age, T1, and T2 values.

Datasets

The study utilized data from 110 female patients who underwent lumbar MAGIC examination and DXA scan at Chongqing Hospital of Traditional Chinese Medicine between November 2020 and April 2021. The patients were categorized into an OP group (56 cases) and a non-OP group (54 cases) based on DXA diagnosis. No new benchmark datasets were introduced.

Key Insights and Ideas

  • MAGIC's Quantitative Power: The MAGIC sequence's ability to simultaneously provide quantitative T1 and T2 maps is a significant advantage, offering intrinsic tissue properties that are more reliable than signal intensity alone.
  • Age-Related Changes: The study confirms that as age increases, BMD and T1 values tend to decrease, while T2 values tend to increase in the lumbar spine. This aligns with the understanding of bone aging and structural changes.
  • T1/T2 as OP Biomarkers: Both T1 and T2 relaxation times demonstrated statistical significance in differentiating between OP and non-OP groups, suggesting their potential as biomarkers for OP.
  • Combined Diagnostic Efficiency: The combination of T1 and T2 values yielded a higher diagnostic efficiency (AUC = 0.985) than either parameter alone, highlighting the synergistic value of these quantitative measures.
  • Predictive Function: The development of a function fitting formula that incorporates age, T1, and T2 to predict BMD offers a promising avenue for direct BMD calculation and OP risk assessment.

Key Results

  • Correlations: Moderate positive correlation was found between T1 and BMD (R=0.636, P<0.001), and a moderate negative correlation between T2 and BMD (R=-0.694, P<0.001). A strong negative correlation was observed between T1 and T2 (R=-0.781, P<0.001).
  • Diagnostic Accuracy: T1 and T2 showed high accuracy in diagnosing OP, with AUC values of 0.982 and 0.978, respectively. The optimal cutoff points were identified as T1 ≤ 0.625s and T2 > 0.095s.
  • Combined Efficacy: The combined use of T1 and T2 achieved an even higher diagnostic efficiency with an AUC of 0.985.
  • Function Fitting: The study derived function fitting formulas for BMD based on age, T1, and T2 for both OP and non-OP groups, with the OP group's formula showing a better fit (lower SSE).

Limitations and Future Work

  • Image Quality at Extremities: The authors noted that while MAGIC generally provides good image quality, there was some impairment in signal-to-noise ratio at the L1 and L5 vertebral bodies, potentially due to respiratory motion artifacts despite saturation zones.
  • Sample Size: While 110 patients is a reasonable sample, larger, multi-center studies would further validate these findings.
  • Age Standardization: The non-OP group was defined as patients <50 years old, which differs from the age-standardized prevalence of OP in Chinese women (60 years old). This might influence the interpretation of results for younger individuals.
  • Future Directions: The paper suggests that MAGIC's ability to provide quantitative parameters and compensate for the limitations of traditional BMD measurements (like not reflecting biomechanical properties) makes it a valuable tool for OP diagnosis and screening. Further research could explore the application of this technique in broader populations and for monitoring treatment efficacy.

Summary updated

BibTeX
@article{chen2023application,
  author = {Yiming Chen and Xiuting Mei and Xuqian Liang and Yi Cao and Cong Peng and Yang Fu and Yulong Zhang and Cuifang Liu and Yang Liu},
  title = {{Application of magnetic resonance image compilation (MAGiC) in the diagnosis of middle-aged and elderly women with osteoporosis}},
  journal = {BMC Medical Imaging},
  year = {2023},
  doi = {10.1186/s12880-023-01010-9},
}
Figure