Intraoperative Ultrasound in Spine Decompression Surgery

A Systematic Review

Jimmy Tat, MD, MSc; Jessica Tat, MD; Samuel Yoon, MD, MSc; Albert J.M. Yee, MD, MSc, FRCSC, DABOS; Jeremie Larouche, MD, FRCSC


Spine. 2022;47(2):E73-E85. 

In This Article


Search Results

Our search strategy yielded 985 of potentially relevant publications (Figure 1). After removal duplicate publications, there were 776 articles that underwent title and abstract screening, 204 articles were screened, and 31 full text articles were reviewed. Studies were only included in qualitative review, as study designs were far too heterogeneous to perform a quantitative synthesis of meta-analysis.

Figure 1.

PRISMA flow diagram.

Types of Study

The studies included in our review consisted of review articles, controlled studies, prospective cohort/longitudinal/follow-up studies. We did not identify any higher tier level of evidence (no randomized controlled trials, nor systematic reviews). Our patient populations included a combination of adults with degenerative spine disease and traumatic fractures of the cervical, thoracic and lumbar spine. We did not identify any articles that applied IOUS in decompression of extradural spine tumors. We excluded studies which utilized ultrasound for localization purposes only, such as those dealing with intradural spine lesions including arachnoid cyst, syringomyelia, hemangioma, spinal arteriovenous malformations, and intradural spinal cord tumors (intramedullary and extramedullary). Many studies did not report clinical outcomes in association with decompression status. We found only one study that reported reliability measures (e.g., interobserver or intraobserver). We used the MINORS tool to assess risk of bias, and overall the studies had a low risk of bias (Table 1, Figure 2).

Figure 2.

Risk of bias across studies assessed using the Methodological Index for Non-Randomized Studies (MINORS) tool.

Ultrasound Technique

We reviewed and summarized the typical protocol for IOUS. Routinely, a posterior decompression is used involving a laminotomy or laminectomy with removal of the ligamentum flavum. The operative field is filled with sterile saline for acoustic coupling. An ultrasound transducer is then placed over the dura to assess the extent of decompression via the laminectomy corridor. A linear array transducer, between 3 and 15 MHz in either straight or hockey stick configuration is commonly used. The probe may be sterilized, or encased in a sterile sheath (Table 2). Images are commonly acquired in the axial and longitudinal views. However, we noted that the location of these ultrasound views were not standardized and often not described.

Cervical Spine

The decompression status using IOUS has been well studied in the cervical spine with prospective case series (Table 3). Spine surgeons from both orthopedic and neurosurgery disciplines have used it in the decompression of cervical myelopathy related to cervical spondolysis, ossification of the posterior longitudinal ligament, and disc herniation.

The definition for decompression was qualitative and referred to the "contact" or "floating" nature of the spinal cord (Figure 3A–C). Adequate decompression was characterized by the ventral aspect of the spinal cord being free floating within the cerebrospinal fluid or absence of contact from the anterior elements (such as OPLL or intervertebral disks). If compressed, the severity of compression was often described based on the resultant shape of the spinal cord. For example, loss of the anterior flat surface of the spinal cord with indentation or bend in the contour. For these definitions, we found conflicting correlations with clinical outcomes. Five out of eleven studies showed improved recovery rates and neurological outcomes with the free floating IOUS finding,[8,9,13–15] whereas another three studies found no difference.[10,11,16] The remainder three studies had no clinical outcomes for comparison.[17–19]

Figure 3.

Intraoperative ultrasound (IOUS) view of the cervical spinal cord collected with an BK5000 US System (BK Medical Inc) and a Hockey Stick X18L5 15Mhz transducer. (A) Sagittal view shows the long axis of the spinal cord. (B) Axial image at the corresponding level, labeled (1) on the sagittal view, demonstrating adequate decompression with "free floating" CSF around the spinal cord. (C) Axial image at the level, labelled (2) on the sagittal view, demonstrating inadequate decompression due to inadequate free floating and contact of the spinal cord with the anterior elements at the level of the spinal canal.

One study attempted to quantify the severity of compression by using IOUS to measure changes in diameter of the spinal cord before and after decompression.[11] This included measurements of spinal cord cross-sectional area, sagittal diameter, and transverse diameter at the level of maximum compression. Cross-sectional area improved significantly from pre-operative to immediately post-decompression as measured by IOUS (37.4 vs. 45.4 mm2, respectively, p < 0.001). However, there were no comparisons made with clinical outcomes and cross-sectional area. The sagittal and transverse diameters were used to calculate a compression ratio (sagittal diameter × 100%/transverse diameter). Using their sample population (n = 44), patients were grouped either above or below the compression ratio sample mean of 28%. A compression ratio ≥28% was regarded as severe spinal cord compression, <28% was a mild compression ratio; however, this had no correlation with clinical recovery rate. There were 22 patients with severe compression (n = 22): n = 14 were categorized as poor clinic recovery (clinical recovery rate <50%), whereas the remaining n = 8 had good clinical recovery (clinical recovery rate >50%).

There were insufficient data available to confirm the reliability of IOUS technique to assess the adequacy of the surgical decompression in the cervical spine. The few studies that have compared IOUS to other imaging modalities including postoperative CT and MRI were inconclusive. Additionally, only one study evaluated interobserver variability in IOUS assessment by surgeons and reported a median kappa coefficient of 0.69 among three independent observers.[8]

Thoracic Spine

In the thoracic spine, decompression status was investigated using prospective and retrospective case series (Table 4 and Table 5). IOUS was applied primarily in the context of posterior decompression of degenerative spine pathology including ossification of the posterior longitudinal ligament, ossification of ligamentum flavum, and disc herniations. It was also helpful in the setting of traumatic fractures of the thoracolumbar spine to assess reduction and adequacy of decompression in fractures leading to spinal stenosis.

Similar to the cervical spine, a theme of decompression status criteria was used in the thoracic spine involving a qualitative definition to restore the ventral epidural space around the spinal cord. This was frequently referred to as "echo-free space" or a "floating" spinal cord, and meant that the spinal cord was free from contact ofthe anterior elements. An inadequate decompression involved contact with anterior elements in all studies. However, likewise, the clinical significance of this definition in the thoracic spine was inconsistent. Only six of 15 studies reported improved clinical recovery with the decompressed status compared to inadequate decompression.[20–25] As for the remainder: two of 13 showed no difference in clinical recovery rates between inadequate and sufficient decompression status[26,27] and five of 13 had no report of clinical outcomes.[3,7,28–32] There were no quantitative definitions for decompression status in the thoracic spine.

The thoracic spine was unique for applying ultrasound in fracture management (Table 5). Six studies used IOUS to guide thoracolumbar fracture reduction by visualizing the spinal canal during ligamentotaxis and if necessary direct repositioning of fracture fragments with instrumentation in real time.[3,7,28–31] Lazennec et al[30] showed that ultrasound was able to identify all main fragments as compared to preoperative CT images (n = 58). Degreif and Wenda[29] found ultrasound could identify 58 of 60 cases with retro-pulsed fragments in the spinal canal, with the 2/60 inconclusive case attributed to severely damaged fragments. IOUS was able to detect cortical fragments with precision of 0.6 mm and 1.2 mm for cancellous bone fragments30. Mean spinal canal stenosis decreased from 60.5% preoperatively to 18.7% after operative reduction monitored by IOUS. Although there was an overall improvement in neurological status pre and post decompression in these patients, there was no relationship made between level of stenosis and clinical outcomes.

The reliability of IOUS technique in the thoracic spine studies was unclear. Few studies included a systematic comparison of the IOUS technique to alternate imaging modality such as CT or MRI.[3,22–24,28–32] Among the ones who performed this comparison, generalized statements such as gross morphological similarities were observed between IOUS spinal canal measurements and pre or post op CT images, without any sort of statistical analysis.

Lumbar Spine

The lumbar spine was least studied area for IOUS guided decompression (Table 6). We identified four studies involving prospective case series with patients suffering from spinal stenosis secondary to disc herniation that were treated with posterior laminectomies.[1,2,33,34]

The criteria for decompression were only qualitative, and nonewerequantitative. Adequate decompression consistently involved a definition that demonstrated restoration of the round shape of the theca sac surrounding the cauda equina. Once the compressive structures were removed, authors reported that the thecal sac would enlarge and appear less echogenic, as it filled with hypoechoic cerebrospinal fluid. We did not identify any clinical outcomes with these definitions.

The lumbar spine was unique for using IOUS to also guide decompression of associated nerve roots. The nerve root could be identified in the axial image as the structure branching from the ventral part of the dural sac and running laterally. Adequate decompression of the nerve root occurred after the diameter was normalized, and the nerve root's course appeared smooth and uncompressed. Likewise, we did not identify any clinical outcomes with these definitions.

The reliability of IOUS technique in the lumbar spine studies remains poorly studied. No studies compared IOUS to other imaging modalities, or examined its interobserver variability.


Pulsatile movements of the spinal cord have been well documented intraoperatively and are thought to be due to changes in pressure in the vascular system of the brain and spinal cord. In our search, we identified five studies that measured pulsatile after decompression (Table 7). Pulsatile motion was used a surrogate measure of extrinsic compression of the cord within the subarachnoid space. The pulsatile movements of the spinal cord were found to be synchronous with arterial pulsations and dependent on neck positioning. Not all patients demonstrated pulsations and there was no association of pulsatily with clinical outcomes.