Search

Study Design: Diffusion tensor magnetic resonance imaging (DT-MRI) reconstruction of lumbar erector spinae (ES) compared with cadaver dissection., Objective: The aim of this study was to reconstruct the human lumbar ES from in vivo DT-MRI measurements and to compare the results with literature and cadaver dissection., Summary of Background Data: DT-MRI enables 3-dimensional in vivo reconstruction of muscle architecture. Insight in ES architecture may improve the understanding of low back function. Furthermore, DT-MRI reconstructions allow individualized biomechanical modeling, which may serve as a clinical tool in injury evaluation and in improvement of understanding of pathologies like scoliosis., Materials and Methods: The lumbar spine of 1 healthy male volunteer was scanned using a 3.0 T clinical MRI scanner. MRI data acquisition consisted of 3 parts: (1) high-resolution T1-weighted turbo spin echo for anatomical reference; (2) DT-MRI measurements for fiber tractography; (3) dual echo gradient echo sequence for signal correction purposes. After processing, DT-MRI data were exported to a custom-built software program for fiber tractography. The resulting reconstructions were anatomically validated by comparison with cadaver dissection and literature., Results: DT-MRI reconstruction of 4 parts of the lumbar ES (thoracic part of iliocostalis lumborum, lumbar part of iliocostalis lumborum, thoracic part of longissimus thoracis, and lumbar part of longissimus thoracis) adequately reflected its complex geometry. Some inaccuracies were found in reconstruction details. DT-MRI reconstructions were generally in agreement with anatomical descriptions from literature and with findings in a dissected cadaver specimen., Conclusions: DT-MRI enables anatomically valid reconstruction of ES architecture. However, for reliable reconstruction of the smallest fascicles and attachments a higher resolution or application of higher-order models is needed. Reconstructions can be used as input for estimation of muscle architecture parameters in individualized biomechanical modeling. Such models are promising as a tool in clinical evaluation and in research of low back pain mechanisms.