Reducing decompression levels by diffusion tensor imaging and conventional magnetic resonance imaging in degenerative lumbar spinal stenosis

Abstract

Background

The selection of a correct level in lumbar spinal stenosis (LSS) remains a common problem and is critically important to the effectiveness of this surgical treatment. Surgery is invasive, and extended laminectomy may lead to secondary surgical complications. The application of diffuse tensor imagining (DTI) and paraspinal mapping (PM) in addition to conventional magnetic resonance imaging (cMRI) may be helpful in this respect. However, the superiority of cMRI + DTI over cMRI+ (DTI or PM) in reducing decompression has not yet been established.

Methods

We compared the surgical levels, determined by cMRI + DTI and cMRI+ (DTI or PM) (self-control). Treatment outcome measurements were performed at two weeks, three months, six months, and twelve months postoperatively.

Results

The surgical levels determined by cMRI ± DTI showed less than that determined by cMRI± (DTI or PM) with statistically significant differences (p value = 0.0199) and cMRI ± PM with no statistically significant differences (p value = 0.5503).

Conclusions

The effectiveness of cMRI ± DTI in the reduction of the surgical levels in degenerative lumbar spinal stenosis is superior than that of cMRI± (DTI or PM).

Keywords:

• Diffuse tensor imagining
• lumbar spinal stenosis
• Oswestry Disability Index
• paraspinal mapping
• visual analog pain scale

Introduction

Anatomically based determination is necessary for the diagnosis of lumbar spinal stenosis (LSS) but is not sufficient to establish the severity of the symptoms.¹ The extent of spinal canal narrowing correlated poorly with symptom severity, and radiologically significant lumbar spinal stenosis was also found in asymptomatic individuals.^1–4 As a consequence, the correlation between symptoms and physical examination findings and the decompression levels, based on common imaging results, is not reliable. The surgical levels, established by conventional magnetic resonance imaging (conventional MRI) and neurogenic examination (NE), may lead to a more extensive surgery and secondary surgical complications in patients without agreement between radiological and clinical symptoms.

Previous experimental studies have showed that DTI may be useful at identifying the source of neural compromise, especially FA. Maybe because that diffusion in the tissue had become more isotropic because of edema, in which fluid is trapped in the tissue, creating an isotropic environment and a reduction in FA. Beaulieu et al.^5,6 reported that Wallerian degeneration after peripheral nerve injury reduces the anisotropy of water diffusion. Several studies indicated that the FA of peripheral nerves was strongly correlated with axonal degeneration and regeneration in rat and mouse sciatic nerves.^7,8 The decrease in the FA values may reflect the degree of microstructural disorganization of the spinal cord, suggesting either local extra-cellular edema or a smaller number of fibers matching a larger extracellular space, or both. On the other hand, minor lesions and edema with roughly preserved fibrillary microstructure of the spinal cord are not associated with major FA changes, as opposed to demyelination, cavitations and necrotic changes.⁹ Thus, the high FA values suggest that the microstructure of the spinal cord is preserved, even in cases with high signal intensity of the spinal cord on T2-weighted images.

The use of diffuse tensor imagining (DTI) and paraspinal mapping (PM) techniques can reduce the decompression levels of lumbar spinal stenosis more than conventional MRI + NE.¹⁰ However, whether cMRI + DTI is superior to cMRI+ (DTI or PM) in the reduction of decompression is still unclear.

Materials and methods

Enrollment

In this study, we included 40- to 90-year-aged symptomatic patients with degenerative lumbar spinal stenosis, detected by cMRI or radiography from October 2017 to October 2019. Since stenosis defining features can be seen on cMRI earlier and clearer than the changes consistent with stenosis can be detected on radiography, patients with degenerative lumbar spinal stenosis on cMRI were eligible for inclusion. The following inclusion criteria were applied: patients with neuroclaudication including both lower back pain and one-leg pain that was consistent with a lumbar spinal stenosis and had persisted for at least six months despite the pharmacologic treatment, physical therapy, or limitation of activity. Leg pain was defined as pain below the buttocks.¹¹ The neurogenic claudication was typical with severe pain or/and disability and a pronounced constriction of the lumbar spinal canal, and thus it was considered for decompression treatment.¹² NE was performed by an experienced spine surgeon, who was blind to the treatment of the patient. The levels of decompression determined by cMRIwere ≥2.

Patients were excluded if they had history of heavy alcohol consumption, history of lower back surgery,^13,14 evidence of polyneuropathy, or technically inadequate cMRI or electromyography (EMG) results.

Interventions

All patients underwent cMRI and symptom, sign, X-ray, PM, and DTI examinations. Then, they were subjected to decompression surgeries (laminectomies or laminotomies) with surgical levels determined by cMRI + DTI. All surgeons were trained, and each of them had been performing at least 50 lumbar spinal decompression surgeries annually. PM and DTI procedures are described below.

Paraspinal mapping (MiniPM)

Typically, the total MiniPM score was used to indicate the extent of paraspinal denervation. In this study, the MiniPM score at each nerve root, a summary of six scores (only a summary of three scores of the fifth needle point-S1 nerve root, represented the score) were found to be associated with the level of the radiculopathy used¹⁵ (Figure 1). Denervation appeared if the paraspinal muscles had fibrillation potentials, positive sharp waves, or complex repetitive discharges¹⁶ (Figure 2). Normal values established in 35 asymptomatic subjects are 0–2 (95% scored ≤2), with a mean of 0.5,¹⁷ our pre-experiment also showed that if the PM scores of the level were ≥ 2 at one side, it was of clinical meaning and the level should be treated surgically. So we set the standard as follow: If a summary of six scores of one level (only a summary of three scores of the fifth needle point [S1 nerve root], represented the score.) was ≥2 one side, and the level should be treated surgically. The PM examination was performed by a qualified electro-diagnostic physician who was blind to the treatment of patients.

Figure 1. The Score sheet: spontaneous activity is scored separately for the first 4 cm insertion (placed in the Score sheet M column) and in the last lcm insertion (placed in the Score sheet S column).²⁵ In this study, L1and S1nerve roots were added to the Score sheet, PM score was the summary of all plus at one level nerve root, one side.

Figure 2. Fibrillation potentials in the denervated muscle. Activity Grades: 1+, fibrillation potentials persistent in at least two areas; 2+, moderate number of persistent fibrillation potentials in three or more areas; 3+, large number of persistent discharges in all areas; 4+, profuse, widespread, persistent discharges filling the baseline.²⁶

cMRI protocol

A 3.0 T cMRI scanner (Achiva; Philips Medical System, Best, The Netherlands) was used in this study. T1-weighted fast spin-echo sequences were obtained using a 453/8.0 ms for TR/TE, 4/0.4-mm section thickness/gap; 176 × 290 matrix; 0.91 × 1.00 × 4.00-mm³ actual voxel size; 0.50 × 0.50 × 4.00 mm³ calculated voxel size, and T2-weighted fast spin-echo (TR/TE, 3604/110) sequences were obtained using 4/0.4-mm section thickness/gap; 176 × 290 matrix; 0.91 × 1.00 × 4.00-mm³ actual voxel size and 0.50 × 0.50 × 4.00 mm³ calculated voxel size.

The following quantitative criteria for central anatomical LSS were used. The dural sac cross-sectional area (DSCSA) ≥ 100 mm² was considered normal; 76 to 100 mm² was considered to be moderately stenotic, and a value ≤76 mm² was classified as severely stenotic. The nerve root compromise in lateral recess was graded as follows: Grade 0, no contact of the disc with the nerve root; Grade 1, contact without deviation; Grade 2, nerve root deviation; Grade 3, nerve root compression. Nerve root compression was considered to be present in case of root deformation.¹⁸ The following criteria for foraminal qualitative assessment were applied: Grade 0, normal foramina with normal dorsolateral border of the intervertebral disk and normal form of the foraminal epidural fat (oval or inverted pear shape); Grade 1, foraminal stenosis and deformity of the epidural fat while the remaining fat was still completely surrounding the exiting nerve root; Grade 2, foraminal stenosis and deformity of the epidural fat with the remaining fat only partially surrounding the exiting nerve root; and Grade 3, stenosis with obliteration of the epidural fat.^18,19 The quantitative measurements were performed by a trained radiologist with blindness to the treatment of patients.

DTI protocol

A 3T cMRI scanner (Achiva; Philips Medical System, Best, The Netherlands) was used in this study. Subjects were scanned in a supine position using an eight-channel phased array spine coil. DTI was performed using an echo-planar imaging sequence with a free-breathing scanning technique. The following imaging parameters were set: 0,600 s/mm² b-value; MPG, 15 directions (Philips DTI medium); 6000/76ms for TR/TE, respectively; axial section orientation, 3/0-mm section thickness/gap; 200 × 200 × 160 mm³ FOV; 64 × 78 matrix; 3.13 × 2.54 × 3.00-mm³ actual voxel size;1.56 × 1.56 × 3.00 mm³ calculated voxel size; NSA:3; 40 total sections; and 5 min 32 s scan time.

T2-weighted 3D fast field echo sequences were obtained using 33/3.9 ms for TR/TE; 80 × 80 matrix; FOV 160 × 160 × 200 mm³; NSA,1; gap, 0 mm; 2.00 × 1.99 × 4.00 mm³ actual voxel size and 0.50 × 0.50 × 2.00 mm³ calculated voxel size

Image analysis

After DTI data were transferred to a PC, and the images were analyzed by Philips Extended Workspace (Philips DICOM Viewer R2.6 SP1). Using the fiber-tracking application software, the anatomical images were superimposed on a fractional anisotropy (FA) map to permit anatomical correlation (Figure 3). The diffusion tensor was calculated through the log-linear fitting method. On axial images, the regions of interest (ROIs) were placed at cauda equina and the nerve roots of a level freehand, to circumscribe cauda equina or nerve roots with minimal inclusion of cerebral spinal fluid (CSF). In the cauda equina, ROIs were placed on the zones the same as the disc, including superior 1/3, middle 1/3, and inferior 1/3 of the disc, taking the minimum value of three zones as the FA value of the cauda equina. In lumbar spinal nerves, ROIs were placed on the ‘intraspinal’, ‘intraforaminal’, and ‘extraforaminal’ zones, taking the minimum value of three zones as the FA value of the nerve root. The FA values were calculated with the software at levels from L1 to S1 in patients. The sizes of ROIs from 25 to 50 mm² and 50 to 150 mm² were selected to be as accurate as possible on the respective nerve roots and the cauda equina to reduce the partial volume effects when the mean FA value was calculated. All DTI analyses were performed twice by two trained radiologists to avoid intra- and interobserver differences,²⁰ and they were blind to the treatment of patients. The following standard was implemented: if a FA value of the lumbar cauda equina or/and the nerve root of the narrow level decreased ≥ 0.1 than that of the non-stenotic or the normal level (commonly took the T12-L1 cauda equina and nerve root values as the reference), it was meaningful and the level should be treated surgically.

Figure 3. The cauda equina MRI T2W image (A) and its FA mapping of DTI (B). ROIs were placed on the cauda equina on the zones equally to the disc plane (superior 1/3, middle 1/3, and inferior 1/3) on the FA mapping and the FA values were calculated (B), the minimum value of three zones was taken as the cauda equina FA at that level; MRI T2W image of bilateral nerve roots (C) and FA mapping of DTI of bilateral nerve roots (D) ROIs were placed on the: ‘intraspinal’, ‘intraforaminal’, and ‘extraforaminal’ zones of bilateral nerve roots on FA mapping and the FA values were calculated (D), the minimum value of three zones was taken as the nerve root FA value; MRI, magnetic resonance imaging ; FA, fractional anisotropy; DTI, diffusion tensor imaging.

Determination of the decompression levels

Determination by cMRI± (PM or DTI)

Based on (1) the central tube ≤76 mm2 or/and foramen or/and lateral recess ≥ Grade 1 narrow determined by cMRI and (2) the score of PM and/or the FA value of DTI was positive, the level was considered for surgical decompression; if the score of PM and the FA value of DTI both were negative, the level was considered for surgery determined only by cMRI. If there were conflicts of the opinion, the two direct of spine surgeons reached a mutual decision through a mutual discussion.

Determined by cMRI ± DTI

Based on (1) the central tube ≤76 mm² or/and foramen or/and lateral recess ≥ Grade 1 narrow determined by cMRI and (2) the FA value of DTI was positive, the level was considered for surgical decompression. If the FA value of DTI were negative, the level was considered for surgery determined only by cMRI. If there were conflicts of the opinion, the two direct of spine surgeons reached a mutual decision through a mutual discussion.

Determined by cMRI ± PM

Based on (1) the central tube ≤76 mm² or/and foramen or/and lateral recess ≥ Grade 1 narrow determined by cMRI and (2) the score for PM was positive, if the score of PM was negative, the level was determined only by cMRI. If there were conflicts of the opinion, the two direct of spine surgeons reached a mutual decision through a mutual discussion.

Determined by cMRI ± NE

A pronounced constriction of the lumbar spinal canal was considered the most important indication for surgical treatment,¹² The mean cross-sectional area of the dural tube at the narrowest level was 68.9 ± 25.7 mm² in the 47 patients with central stenosis²¹ and based on the experience of our spine specialists that decompression levels generally are within the scope of the central tube ≤76 mm² or/and foramen or/and lateral recess ≥ Grade 1 narrow, we set the standard as follows:

Levels of decompression were including the level of the central tube ≤76 mm² or/and foramen or/and lateral recess ≥ Grade 1 narrow determined by cMRI and that located by NE in term of the American association of spinal cord injury (ASIA).if NE cannot locate the level, determined only by cMRI. If there were conflicts of the opinion, the two direct of spine surgeons reached a mutual decision through a mutual discussion.

Outcomes

Our primary outcomes were the averages of the reference FA value, positive, negative FA value and positive, negative PM score, as well as the levels. The reference FA value was referred to the FA value of L1, whereas the positive and the negative FA value referred to the FA values of the positive and negative level, respectively. Each level included the cauda equina and the nerve roots on the two sides. The positive and negative PM scores denoted the PM scores of the positive and negative levels, and each level included the nerve roots on the two sides. The decompression levels were determined by cMRI± (PM or DTI), cMRI ± DTI, cMRI ± PM and cMRI ± NE.

The visual analog scale pain scores for both back and leg symptoms (VAS-BP, VAS-LP) and the Oswestry Disability Index (ODI) have been used as secondary outcomes (on a scale from 0 to100) in patients with lumbar spinal stenosis.²² All patients were blind to pain scores and ODI.

Assessments

The primary outcomes were assessed preoperatively. The secondary outcomes were evaluated preoperatively and two weeks, three months, six months, and twelve months postoperatively. Postoperative assessments were employed to determine the trajectory and stability of the treatment responses. Institutional Ethics Review Board approval was obtained before commencing data collection.

Statistical analysis

All measurement values were expressed as mean ± standard deviation. Preoperatively and the two weeks postoperatively, we conducted pairwise comparison of Receiver Operating Characteristics (ROC) curves of surgical levels of cMRI± (PM or DTI), cMRI ± DTI, cMRI ± PM and cMRI ± NE and primary analysis of the covariance of VAS-BP, VAS-LP, and ODI. The difference between VAS-BP, VAS-LP, and ODI preoperatively and two weeks postoperatively were assessed by t-test. All statistical analyses were performed using IBM SPSS version 19.

Results

Characteristics of the study population

From October 2017 to October 2019, 50 patients with degenerative lumbar spinal stenosis, detected on conventional MRI, were eligible.

Outcomes

All reference FA values of 50 patients were obtained at the L1 level (Table 1). The average reference FA values and the levels cauda equina, 0.437 ± 0.028 (50); left nerve root, 0.458 ± 0.027 (50); right nerve root, 0.468 ± 0.027 (50).The average of the FA values of positive levels and the levels were cauda equina, 0.294 ± 0.034 (89); left nerve root, 0.313 ± 0.038 (9); right nerve root, 0.309 ± 0.040 (16). The average of the FA values of negative levels and the levels were cauda equina, 0.406 ± 0.046 (66); left nerve root, 0.490 ± 0.074 (66); right nerve root, 0.488 ± 0.058 (66) .No statistically significant difference was found between FA values established by two radiologists. The average values of the PM scores of the positive levels and the levels were left nerve root, 3.35 ± 1.38(49); right nerve root, 2.98 ± 1.32(46).The average of the PM scores of negative levels and the levels were left nerve root, 1(22); right nerve root, 1(26) (Table 2). All lumbar spinal levels were with respect to the decompression levels determined by only cMRI except for the reference levels.

Table 1. Averages of FA values of levels and the levels.

Reference FA values^a (n)			Positive FA values (n)			Negative FA values (n)
Cauda equinas (n)	Left nerve roots (n)	Right nerve roots (n)	Cauda equinas (n)	Left nerve roots (n)	Right nerve roots (n)	Cauda equinas (n)	Left nerve roots (n)	Right nerve roots (n)
0.437 ± 0.028 (50)	0.458 ± 0.027 (50)	0.468 ± 0.027 (50)	0.294 ± 0.034 (89)	0.313 ± 0.038 (9)	0.309 ± 0.040 (16)	0.406 ± 0.046 (66)	0.490 ± 0.074 (66)	0.488 ± 0.058 (66)

^aAll levels were considered to be surgical decompression determined by cMRI.

FA: fractional anisotropy; cMRI: conventional magnetic resonance imaging.

Table 2. Levels and scores of PM.

Project	Positive PM scores		Negtive^a PM scores
	Left nerve roots	Right nerve roots	Left nerve roots	Right nerve roots
Levels (n)	49	46	22	26
Averages (scores one level)	3.35 ± 1.38	2.98 ± 1.32	1	1

PM: paraspinal mapping.

^aPM scores＝1, one level, one side.

The surgical levels determined by cMRI ± PM, cMRI ± DTI, and cMRI ± (DTI or PM) were all less than that determined by cMRI ± NE with statistically significance (p value ≤ 0.0001, p value = 0.0001, p value = 0.0001, respectively). The surgical levels determined by cMRI ± DTI showed less than that determined by cMRI± (DTI or PM) with statistically significant differences (p value = 0.0199) and cMRI ± PM with no statistically significant differences (p value = 0.5503; Table 3). ROC curves of four methods showed that area under curve (AUC) of MRIDTI was more than MRIPM, MRIPM was more than MRI (PM or DTI), while MRI (PM or DTI) was more than MRINE. Therefore, surgical levels determined by cMRI ± DTI were most less (Table 4, Figure 4). No conflicts of the opinions happened in the decision of the decompression levels with all patients.

Figure 4. ROC curves of Four methods showed that area under curve (AUC) of cMRIDTI was more than cMRIPM, cMRIPM was more than cMRI (PM or DTI), while cMRI (PM or DTI) was more than cMRINE. cMRI, conventional magnetic resonance imaging; DTI, diffuse tensor imaging; PM, paraspinal mapping; NE, neurogenic examination.

Table 3. Levels of decompression determined by cMRI + NE, cMRI + PM, cMRI+(DTI or PM), and cMRI + DTI.

Methods	cMRI + NE	cMRI + PM	cMRI + DTI	cMRI+(DTI or PM)
Decompression levels	3.20 ± 0.79	2.03 ± 0.55	1.96 ± 0.73	2.04 ± 0.57

cMRI: conventional magnetic resonance imaging; DTI: diffuse tensor imaging; PM: paraspinal mapping; NE: neurogenic examination.

Table 4. Pairwise comparison of ROC curves.

Subjects	MRIDTI-MRINE	MRIDTI-MRIPM	MRIDTI-MRI (PM or DTI)	MRINE-MRIPM	MRINE-MRI (PM or DTI)	MRIPM-MRI (PM or DTI)
p values	p < 0.0001	p = 0.5503	p = 0.0199	p < 0.0001	p < 0.0001	p = 0.3522

cMRI: conventional magnetic resonance imaging; DTI: diffuse tensor imaging; PM: paraspinal mapping; NE: neurogenic examination.

Two weeks postoperatively, all postoperative ODI, VAS-BP, and VAS-LP scores were measured and found to be lower than the preoperative ODI, VAS-BP, and VAS-LP scores with statistically significant differences (p value = 0.000, p value = 0.000, and p value = 0.000, respectively), and the average of ODI, VAS-BP, and VAS-LP scores were greatly reduced from (43.78 ± 7.41, 99.80 ± 22.48, 106.10 ± 18.76, correspondingly) before surgeries to (16.14 ± 2.12, 22.04 ± 6.59, 22.98 ± 6.71, respectively) two weeks after the surgeries (Figure 5, Table 5). Time-order improvements were observed in the VAS-BP, VAS-LP scores and ODI scores at two weeks, three months, six months, and twelve months postoperatively (Table 5).

Figure 5. Improvements in functional and pain scores at 2 weeke after surgery compared to that before surgery. ODI, Oswestry Disability Index; VASBP, visual analog scale for back pain; VASLP, visual analog scale for leg pain.

Table 5. Scores.

Projects	Preoperative	2 weeks	3 months	6 months	12 months
VAS-BP	99.80 ± 22.48	22.04 ± 6.59	16.96 ± 3.19	13.86 ± 2.76	9.38 ± 1.55
VAS-LP	106.10 ± 18.76	22.98 ± 6.71	16.96 ± 3.08	13.84 ± 2.79	8.62 ± 1.71
ODI	43.78 ± 7.41	16.14 ± 2.12	13.80 ± 1.63	11.80 ± 1.51	7.24 ± 1.42

ODI: Oswestry Disability Index; VASBP: visual analog scale for back pain; VASLP: visual analog scale for leg pain.

Discussion

Our FA values of the nerve root were not comparable to those obtained in the study of lumbar spinal nerves conducted by Balbi et al.²³ (0.218), vanderJagt et al.²⁴ (0.31), and Budzik et al.²⁵, might because of the different software calculation methods. Our reference FA values and negative FA values of the cauda equina were larger than those of the gray matter (0.32 ± 0.11), less than white matter (0.63 ± 0.08),^26,27 and lower than the average of the cauda equina (0.492),²⁸ because at the L1 level, the FA value we measured was actually a FA value of mixture of gray matter, white matter, and cerebrospinal fluid, and at L2 – S1 levels, the FA value we measured was actually a FA value of the mixture of cauda equina nerve and cerebrospinal fluid, the lower FA value of cerebrospinal fluid would reduce the FA value. This affected FA values of the reference and narrow level, but not their difference.

The PM scores of positive compared with the negative PM scores obviously increased with significant differences (left nerve root, p = 0.000; right nerve root, p = 0.000) showed that the standard of PM was statistically significant. A positive EMG, based on spontaneous activity findings, can reassure clinicians that a lesion seen on an imaging study is indeed a pain generator.²⁹ Haig et al.³⁰ argued that imaging does not differentiate between symptomatic from asymptomatic individuals, whereas electrodiagnosis does. They believe that radiographic findings alone are insufficient to justify treatment for spinal stenosis. Normal values established in 35 asymptomatic subjects are 0–2 (95% scored ≤2), with a mean of 0.5,¹⁷ our pre-experiment also showed that if the PM scores of the level was ≥2 at one side, it was of clinical meaning and the level should be treated surgically. So we set the standard as follow: If a summary of six scores of one level (only a summary of three scores of the fifth needle point [S1 nerve root], represented the score.) was ≥2 one side, and the level should be treated surgically.

The surgical levels determined by cMRI ± DTI showed less than that determined by cMRI± (DTI or PM) with statistically significant differences and cMRI ± PM with no statistically significant differences, which indicated that cMRI ± DTI was superior to cMRI± (DTI or PM) in the reduction of the surgical level in degenerative lumbar spinal stenosis.

We measured all postoperative ODI, VAS-BP, and VAS-LP scores two weeks postoperatively. Improvements were established in comparison with the preoperative ODI, VAS-BP, and VAS-LP scores with statistically significant differences. Furthermore, the average ODI, VAS-BP, and VAS-LP scores were considerably lower, indicating that good results were achieved by the surgical treatment. In the postoperative follow-up, there were two patients with lumbar compression fractures (because of osteoporosis), one patient with femoral neck fracture, which were not associated with degenerative lumbar spinal stenosis. The postoperative VAS-BP, VAS-LP, and ODI scores in some cases had fluctuations, but the averages exhibited a trend toward timely improvement; and none of the patient symptoms recurred or/and exacerbated or/and required a repeat surgery.

Limitations

Some limitations to our study should be acknowledged. First, we investigated a small number of subjects, and the follow-up period was limited. Thus, further studies are needed to investigate the validity of our findings in larger populations with longer follow-up periods. Second, we could not repeat DTI and PM postoperatively because of spinal instrumentation artifacts, including those from the pedicle screw systems and surgical scar.

Conclusions

The findings of the present prospective study showed that the efficacy of conventional MRI ± DTI is superior to that of conventional MRI± (DTI or PM) in the reduction of surgical levels in degenerative lumbar spinal stenosis.

Ethical approval

This study was conducted in accordance with the ethical standards in the 1964 Declaration of Helsinki and relevant regulations of the US Health Insurance Portability and Accountability Act (HIPAA).

Author contributions

Hua-Biao Chen and Yu-Liang Huang conceived of the study, carried out the patient collection, participated in the design of the study, performed the statistical analysis and drafted the manuscript. Min-Chen carried out the patient collection, participated in the design of the study and carried out the radiological and paraspinal mapping examination. Hong-Bo Wu conceived and guied the study design and experiment. All authors read and approved the final manuscript.

Acknowledgement

This paper had get the help from “Huizhou city clinical key specialty project funding”.

Disclosure statement

No potential conflict of interest was reported by the author(s).