Subjects
The DLS group comprised 40 patients with DLS diagnosed in Tianjin Hospital from March 2021 to June 2022. The inclusion criteria for this group were no history of scoliosis and a diagnosis of lumbar scoliosis with a Cobb angle of >10° by examination; an age >50 years; routine imaging examinations, including a standing full-length spine film and lumbar spine MRI; no nerve root lesions; and no history of treatment. The control group comprised 40 healthy individuals who received physical examinations in Tianjin Hospital from March 2021 to June 2022. The inclusion criteria for the control group were no scoliosis, an age of >50 years, a standing full-length spine film and lumbar spine MRI, no radiculopathy, and no history of treatment.
The exclusion criteria for both groups were significant limitation of standing or lumbar extension and flexion; abnormal vertebral development, lumbar fracture, lumbar spondylolisthesis, or similar disorders, and a history of an operation on the lumbar spine, pelvis, or lower limb.
There were no significant differences in sex, age, height, weight, or body mass index between the DLS and control groups (Table 1).
Table 1 Demographic characteristics of the control group and DLS group (mean ± SD).
Variable
|
Control(N=40)
|
DLS(N=40)
|
Statistcs
|
P value
|
Gender(F/M)
|
24/16
|
28/12
|
χ2=3.31
|
P=0.143
|
Age(years)
|
62.68±7.97
|
64.9±8.05
|
t=-0.543
|
P=0.423
|
Height(m)
|
1.65±0.04
|
1.64±0.04
|
t=-1.841
|
P=0.217
|
Weight(kg)
|
72.17±9.22
|
69.23.0±8.25
|
t=-1.364
|
P=0.273
|
BMI
|
25.93(16.27-29.39)a
|
25.33±3.41
|
Z=-1.217
|
P=0.321
|
aThe BMI of the control group showed non-normal distribution, and the median was used to describe the data.
Imaging conditions
All subjects underwent radiography and MRI examinations using the same equipment. Lumbar X-ray examinations were performed using a 500-mA digital radiography system (AXIOM Aristos VX Plus; Siemens Healthineers, Erlangen, Germany) with 75 kV in the anteroposterior position and 85 kV in the lateral position using an automatic exposure control system. High-resolution 1.5T MRI (MAGNETOM Essenza; Siemens Healthineers) was used for the lumbar spine with a slice width of 4 mm, acquisition matrix of 256 × 256, T1-weighted imaging repetition time of 587 ms (echo time of 11 ms), and T2-weighted imaging repetition time of 3000 ms (echo time of 93 ms).
Imaging evaluation
In the DLS group, the Cobb angle between the superior endplate of the UEV and the inferior endplate of the LEV was measured on the coronal plane of the lumbar spine. Thirty patients in the DLS group had left convex scoliosis and 10 patients had right convex scoliosis, and the mean coronal Cobb angle was 16.81° ± 6.24° (Fig. 1A).
T2 sequence images were selected for measurement, and three T2-weighted imaging images were taken at each disc level from T12/L1 to L5/S1, and intermediate slices were selected as the research object. Using ImageJ software v2.1.4.7 (National Institutes of Health, Bethesda, MD, USA), the region of interest was determined by manually marking the muscle boundaries. The inferior disc levels of the UEV, AV, and LEV were selected for measurement in the DLS group, and the inferior disc levels of the L1, L3, and L5 vertebrae were selected in the control group.
To reduce the effect of body size, height, and weight on muscle parameters, the ratio of the cross-sectional area (CSA) to the same-level disc CSA was used as the relative CSA (rCSA) [9]. The fat tissue in the 8-bit image was shown in red using the threshold technique [7] (Fig. 1B). The red area in the CSA of the MF was measured as the fat infiltrated area (FIA), and the percentage of FIA in the CSA of the MF was calculated as the percentage of FIA (%FIA) [4].
To facilitate comparison of the imaging parameters (rCSA and %FIA) of the MF on the convex and concave sides, the CSA difference index (CDI) and FIA difference index (FDI) were calculated. The following formulas were used: CDI = rCSA on the concave side − rCSA on the convex side, and FDI = %FIA on the concave side − %FIA on the convex side. The CDI and FDI are considered to represent the asymmetric degree of MF degeneration on the convex and concave sides [8,10]. The DLS group showed obvious asymmetry in the degeneration of the MF on the convex and concave sides (Fig. 2).
Reliability of the method
Muscle measurements on MRI were independently assessed by two examiners. Intraobserver repeatability and interobserver reliability were calculated by one-way analysis of variance and calculation of the intraclass correlation coefficient. According to the Fleiss criterion [11], an intraclass correlation coefficient of less than ±0.40 indicates poor reliability, ±0.40 to 0.75 indicates fair or good reliability, and ±0.75 to 1.00 indicates excellent reliability. All data were measured with excellent intraobserver and interobserver reliability (Table 2).
Table 2 The correlation coefficients for intraobserver repeatability and interobserver reliability for both parameters were >0.9, indicating that the measurements were reliable.
Parameter
|
Level
|
intraobserver
|
interobserver
|
|
UEV
|
0.962
|
0.981
|
%FIA
|
AV
|
0.973
|
0.967
|
|
LEV
|
0.992
|
0.937
|
|
UEV
|
0.977
|
0.964
|
rCSA
|
AV
|
0.984
|
0.979
|
|
LEV
|
0.974
|
0.980
|
sEMG measurements
An Ultium sEMG system (Noraxon USA, Scottsdale, AZ, USA) was used to record sEMG signals at a sampling rate of 2000 Hz. According to the European sEMG recommended standards [12], after skin preparation, Ag/AgCl bipolar circular surface electrodes (Shanghai, China) (44 × 22 mm; electrode spacing, 18 mm) prepared from conductive gel were placed on the bilateral MF muscle bellies at the UEV, AV, and LEV levels in the DLS group and at the L1, L3, and L5 levels in the control group(Fig. 3), consistent with the direction of the muscle fibers [13,14].
Tasks and procedures
The movement tasks included standing still, maximum standing flexion, and maximum standing extension. The subjects were asked to stand for 5 seconds in the resting state and then maintain maximum flexion and maximum extension for 5 seconds, and the sEMG signals in the three movement tasks were analyzed. Each subject was tested three times, and the average result of the three tests was used for further analysis (Fig. 4).
Before the formal measurement, the subjects were instructed to perform lumbar extension and flexion activities to familiarize themselves with the experimental process, avoiding participation of the pelvis and lower limbs as much as possible. To prevent fatigue, the patients had a 1-minute rest period between any two consecutive tests.
Inertial measurement unit (IMU)
The three sEMG sensors on the left side were integrated with the IMU [15,16] and fixed on the spinous processes at the L1, L3, and L5 levels using double-sided tape to collect motion signals (Fig. 3,5).
sEMG signal processing
The raw sEMG signal was processed using a Butterworth bandpass filter (30–500 Hz), and baseline noise was removed. Using the kinematic data acquired simultaneously by the IMU, the 3-second sEMG data with smooth movements were selected for analysis. The root mean square was calculated using MATLAB software (MATLAB R2020a; MathWorks, Natick, MA, USA) and normalized to the ratio to the maximal voluntary isometric contraction (MVIC), or %MVIC, representing the degree of muscle activation. Larger values indicate higher levels of muscle activation, allowing comparison of different levels of muscle activity [17].
Statistical analysis
Statistical analysis was performed using SPSS 25.0 (IBM Corporation, Armonk, NY, USA). Measurement data are presented as mean ± standard deviation. All data were analyzed by descriptive analysis and normality tests. Non-parametric tests were used to compare the demographic data of the two groups. Multifactorial analysis of variance was used to evaluate the effects of influencing factors (experimental grouping, left and right/convex and convex, and different levels of scoliosis) on imaging parameters and sEMG activation. Pearson’s correlation test was used to analyze the relationship between the coronal Cobb angle and the radiographic and sEMG data differences (degenerative asymmetry). A P value of <0.05 was considered to indicate a statistically significant difference.