Dual-source ultrasound identification of the lumbar epidural space allows real-time guidance of midline lumbar epidural needle insertion in a porcine model
Abstract Number: 65
Abstract Type: Original Research
Introduction. Epidural insertion in the lumbar region can be technically challenging in patients without easily discernable surface anatomical landmarks. While ultrasound has been described for pre-puncture identification of spinal anatomy, real-time guidance of epidural needle insertion in the lumbar region has been limited to the paramedian approach. The steep angle required for in plane needle visualization is a significant limitation of this technique.
This study demonstrates that the lumbar spinal anatomy and the epidural space can be visualized from a midline re-slice of a three dimensional (3D) ultrasound volume and that using a needle guidance apparatus affixed to the ultrasound probe, real-time guidance of an epidural needle insertion via the midline approach can be achieved.
Methods. A customized dual-probe apparatus holding an ultrasound probe on either side of the posterior spine was designed. A two dimensional (2D) probe was positioned to provide a real-time transverse paramedian view of the lumbar interspace and a 3D probe was positioned on the contralateral side to provide a re-sliced plane orthogonal to the transverse 2D probe and calibrated to provide a midline view. The needle path was visualized in this midline re-sliced view and was guided by a needle holder affixed to the ultrasound probe.
Experiments were conducted on two freshly slaughtered porcine subjects and all lumbar vertebral interspaces were identified and instrumented. The depth to the epidural space was measured in the transverse paramedian view, the midline sagittal re-sliced view, and an additional midline transverse view that has been previously described . The depth to the epidural space was further determined by a standard loss of resistance technique using a 17 gauge Tuohy needle guided by the needle holder.
Results. Using dual-source ultrasound, all lumbar interspaces and respective ligamenta flava were identified. Epidural needle insertion was visualized in real-time in either the transverse paramedian view or the midline sagittal plane of the three dimensional volume. Measured depth to the epidural space on 3D ultrasound differed from actual depth as determined by loss of resistance by a mean of 5.4 mm (SD 6.7 mm). Bland-Altman 95% limits of agreement were -9.3 to 20.1 mm.
Discussion. This preliminary study demonstrates the feasibility of a midline lumbar epidural insertion guided in real-time by a dual-source ultrasound apparatus providing a midline sagittal view and a transverse paramedian view. The dual-source ultrasound was able to visualize the epidural space and determine the depth to this space. The needle path could be visualized in the midline sagittal plane and guided by a fixed needle holder. This represents a novel approach to visualizing the midline spine in the lumbar region that may potentially allow single-operator real-time ultrasound guided epidural needle insertion.
 Anesth Analg 2007; 104: 1188-92