Automatic feature extraction and fusion recognition of motor imagery EEG using multilevel multiscale CNN.

A motor imagery EEG (MI-EEG) signal is often selected as the driving signal in an active brain computer interface (BCI) system, and it has been a popular field to recognize MI-EEG images via convolutional neural network (CNN), which poses a potential problem for maintaining the integrity of the time-frequency-space information in MI-EEG images and exploring the feature fusion mechanism in the CNN. However, information is excessively compressed in the present MI-EEG image, and the sequential CNN is unfavorable for the comprehensive utilization of local features. In this paper, a multidimensional MI-EEG imaging method is proposed, which is based on time-frequency analysis and the Clough-Tocher (CT) interpolation algorithm. The time-frequency matrix of each electrode is generated via continuous wavelet transform (WT), and the relevant section of frequency is extracted and divided into nine submatrices, the longitudinal sums and lengths of which are calculated along the directions of frequency and time successively to produce a 3 × 3 feature matrix for each electrode. Then, feature matrix of each electrode is interpolated to coincide with their corresponding coordinates, thereby yielding a WT-based multidimensional image, called WTMI. Meanwhile, a multilevel and multiscale feature fusion convolutional neural network (MLMSFFCNN) is designed for WTMI, which has dense information, low signal-to-noise ratio, and strong spatial distribution. Extensive experiments are conducted on the BCI Competition IV 2a and 2b datasets, and accuracies of 92.95% and 97.03% are yielded based on 10-fold cross-validation, respectively, which exceed those of the state-of-the-art imaging methods. The kappa values and p values demonstrate that our method has lower class skew and error costs. The experimental results demonstrate that WTMI can fully represent the time-frequency-space features of MI-EEG and that MLMSFFCNN is beneficial for improving the collection of multiscale features and the fusion recognition of general and abstract features for WTMI.

Complications and Prevention Strategies of Oblique Lateral Interbody Fusion Technique.

OBJECTIVE: To analyze the early complications and causes of oblique lateral interbody fusion, and put forward preventive measures. METHODS: There were 235 patients (79 males and 156 females) analyzed in our study from October 2014 to May 2017. The average age was 61.9 ± 0.21 years (from 32 to 83 years). Ninety-one cases were treated with oblique lateral interbody fusion (OLIF) alone (OLIF alone group) and 144 with OLIF combined with posterior pedicle screw fixation through the intermuscular space approach (OLIF combined group). In addition, 137/144 cases in the combined group were primarily treated by posterior pedicle screw fixation, while the treatments were postponed in 7 cases. There were 190 cases of single fusion segments, 11 of 2 segments, 21 of 3 segments, and 13 of 4 segments. Intraoperative and postoperative complications were observed. RESULTS: Average follow-up time was 15.6 ± 7.5 months (ranged from 6 to 36 months). Five cases were lost to follow-up (2 cases from the OLIF alone group and 3 cases from the OLIF combined group). There were 7 cases of vascular injury, 22 cases of endplate damage, 2 cases of vertebral body fracture, 11 cases of nerve injury, 18 cases of cage sedimentation or cage transverse shifting, 3 cases of iliac crest pain, 1 case of right psoas major hematoma, 2 cases of incomplete ileus, 1 case of acute heart failure, 1 case of cerebral infarction, 3 case of left lower abdominal pain, 9 cases of transient psoas weakness, 3 cases of transient quadriceps weakness, and 8 cases of reoperation. The complication incidence was 32.34%. Thirty-three cases occurred in the OLIF alone group, with a rate of 36.26%, and 43 cases in the group of OLIF combined posterior pedicle screw fixation, with a rate of 29.86%. Fifty-seven cases occurred in single-segment fusion, with a rate of 30.0% (57/190), 4 cases occurred in two-segment fusion, with a rate of 36.36% (4/11), 9 cases occurred in three-segment fusion, with a rate of 42.86% (9/21), and 6 cases occurred in four-segment fusion, with a rate of 46.15% (6/13). CONCLUSION: In summary, OLIF is a relatively safe and very effective technique for minimally invasive lumbar fusion. Nonetheless, it should be noted that OLIF carries the risk of complications, especially in the early stage of development.

[Case control study on two different surgical approaches combined fixation with lumbar interbody fusion for the treatment of single segmental lumbar vertebra diseases].

OBJECTIVE: To discuss the advantages and disadvantages of two different surgical approaches combined fixation with lumbar interbody fusion in treating single segmental lumbar vertebra diseases. METHODS: The clinical data of 86 patients with single segmental lumbar vertebra diseases treated from June 2011 to June 2013 was retrospectively analyzed. There were 33 males and 53 females, aged from 28 to 76 years old with an average of 53.0 years. Among them, there were 39 cases of lumbar disc degeneration, 22 cases of lumbar disc herniation complicated with spinal canal stenosis, 9 cases of huge lumbar disc herniation and 16 cases of lumbar degenerative spondylolisthesis (Meyerding degree I ). Lesion sites contained L3, 4 in 5 cases, L4, 5 in 70 cases and L5S1 in 11 cases. All the patients were treated with internal fixation and lumbar interbody fusion with 45 cases by midline incision approach (median incision group) and the other 41 cases by channel-assisted by muscle-splitting approach(channel group). Incision length, operation time, intraoperative bleeding and postoperative drainage were recorded in two groups. Visual analogue scale(VAS) was used to assess lumbar incision pain 72 h after operation. Depended on imaging results to compare the changes of the disc space height in lesion in preoperative, postoperative and final follow-up, the coronal and sagittal Cobb angle in preoperative and final follow-up, the area of multifidus and the degree of multifidus fat deposition before and after operation between two groups. Loosening or fragmentation of internal fixation, displacement of intervertebral cage and interbody fusion were observed in each group. Japanese Orthopedic Association (JOA) scoring system was used to evaluate the function before operation and at the final follow-up. RESULTS: The channel group was superior to the median incision group in incision length and postoperative drainage while the median incision group was less than the channel group in the operation time and intraoperative bleeding. The average VAS score of lumbar incision 72 h after operation was 1.50 points in median incision group and 0.97 points in channel group, and there was significant difference between two groups(P<0.05). No incision infection was found, but there were 4 cases of incisional epidermal necrosis, 1 case of incision healed badness, and 3 cases of nerve injury in channel group. The incidence of cacothesis of pedicle screw were 5.0% and 3.6% in median incision group and channel group respectively, and there was no significant difference between two groups(P>0.05). The incidence of cacothesis of translaminar facet screw were 6.6% and 12.2% in median incision group and channel group respectively, and there was significant difference between two groups(P<0.05). All the patients were followed up for 12 to 36 months with a mean of 22.8 months. The changes of disc space height had statistical difference between preoperative and postoperative(P<0.05) in all patients, but there was no significant difference between postoperative and final follow-up(P>0.05), however, there was no significant difference 3 days after operation and final follow-up between two groups(P>0.05). At final follow-up, coronal and sagittal Cobb angle were obviously improved in all patients(P<0.05), but there was no significant difference between two groups(P>0.05). One year after operation, the area of multifidus in median incision group was (789.00±143.15) mm² less than preoperative(1 066.00±173.55) mm² (P<0.05), and in channel group, was(992.00±156.75) mm² at 1 year after operation and(1 063.00±172.13) mm² preoperatively, there was no significant difference between them(P>0.05), however, there was significant difference one year after operation between two groups (P<0.05) . About the degree of multifidus fat deposition, there was significant difference between one year after operation and preoperation in median incision group (P<0.05), but there was no significant difference between one year after operation and preoperation in channel group (P>0.05), and there was significant difference at one year after operation between two groups(P<0.05). During the follow-up period, neither pedicle screw and/or translaminar facet screw loosening, displacement or fragmentation nor displacement of intervertebral cage were found. The lumbar interbody fusion rate was 95.6% in median incision group and was 95.1% in channel group, and there was no significant difference between two groups(P>0.05). No obvious adjacent segmental degeneration was observed in fixed position. JOA score in median incision group was significantly increased from 8-16 points (average: 12.77±2.56) preoperative to 21-29 points (average: 25.20±2.43) at final follow-up(P<0.05); and in channel group was significantly increased from 8-16 points (average: 12.64±2.37) preoperative to 23-29 points(average: 26.7±1.82) at final follow-up(P<0.05); there was also significant difference between two groups at final follow-up. CONCLUSIONS: Compared to the median incision approach, unilateral pedicle screw combined with contralateral translaminar facet screw fixation using channel-assisted by muscle-splitting approach has advantages of small incision, less trauma, fast recovery and so on. However, it also has shortages such as high surgical complications incidence, especially in cases that.

[Two different fixation methods combined with lumbar interbody fusion for the treatment of two-level lumbar vertebra diseases: a clinical comparison study].

OBJECTIVE: To investigate the advantages and disadvantages of unilateral pedicle screw fixation combined with contralateral translaminar facet screw fixation and interbody fusion with cages in the treatment of two-level lumbar vertebra diseases, by comparing bilateral pedicle screw fixation and interbody fusion with cages. METHODS: Forty-nine patients with two-level lumbar diseases who received treatments from June 2009 to December 2011 were included in this study. Among these patients, 23 patients received unilateral pedicle screw fixation combined with contralateral translaminar facet screw fixation and interbody fusion with cages (combined fixation group) and the remaining 26 patients underwent bilateral pedicle screw fixation and interbody fusion with cages (bilateral fixation group). These patients consisted of 17 males and 32 females, ranging in age from 29 to 68 years old. Among these patients, lumbar intervertebral disc herniation accompanied by the spinal canal stenosis was found in 29 patients, degenerative lumbar disc diseases in 17 patients and lumbar degenerative spondylolisthesis (degree I) in 3 patients. The lesions occurred at L2,3 and L3,4 segments in 1 patient, at L3,4 and L4,5 segments in 30 patients, and at L4,5 segment and L5S1 segment in 18 patients. Wound length, operation time, intraoperative blood loss and postoperative wound drainage were compared between two groups. Intervertebral space height in the lesioned segment before and during surgery and at the latest follow up was also compared between two groups. Before surgery and at the latest follow-up, the Cobb angle of the coronal plane and sagittal plane of the lumbar spine, loosening or breakage of internal fixations, the dislocation of intervertebral cages, and interbody fusion were all evaluated in each group. The visual analogue scale (VAS) was used to measure lumbar incision pain. The Japanese Orthopedic Association (JOA) scoring system was used to evaluate the function before surgery and at the latest follow-up. RESULTS: No wound infection or skin necrosis was observed after surgery in all patients. No cerebrospinal fluid leakage, nerve root injury, cauda equia injury or worsened neural function in the lower limb occurred in all patients during and after surgery. Wound length, operation time, intraoperative blood loss and postoperative wound drainage in the combined fixation group were superior to those in the bilateral fixation group. At postoperative 72 hours, the VAS score in the combined fixation group (1 to 4 points, mean 2.35±1.20) was significantly lower than that in the bilateral fixation group (2 to 5 points, mean 3.11±1.00; P<0.05). All the patients were followed up for 12 to 48 months, with a mean of 29 months. After surgery, intervertebral space height was well recovered in each patient and it was well maintained at the latest follow-up, and there was no significant difference between two groups (P>0.05). During follow-up, pedicle screw and translaminar facet screw loosening, dislocation or breakage and dislocation of intervertebral cages were all not found. At the latest follow-up, the Cobb angle of the coronal plane and sagittal plane of the lumbar spine was obviously improved and was not significantly different between two groups (P>0.05). The lumbar interbody fusion rate was 93.5% and 96.2% in the combined fixation group and bilateral fixation group, respectively, and there was no significant difference between them (P>0.05). There was a significant difference in JOA score between before surgery and at the latest follow-up in each patient (P<0.05), and at the latest follow-up, significant difference in JOA score was found between two groups (P<0.05). CONCLUSION: Compared to bilateral pedicle screw fixation and lumbar interbody fusion with cages, unilateral pedicle screw fixation combined with contralateral translaminar facet screw fixation and lumbar interbody fusion with cages shows advantages including small skin incision, minimal invasion, ease of operation, highly reliable stability, high interbody fusion rate, rapid recovery in the treatment of two-level lumbar vertebra diseases and therefore can be preferred as a treatment method of this disease.

Mixed Fixation and Interbody Fusion for Treatment Single-Segment Lower Lumbar Vertebral Disease: Midterm Follow-up Results.

OBJECTIVE: To investigate the midterm outcomes of unilateral pedicle screws combined with contralateral gunsight-guided translaminar facet screw fixation and interbody fusion for treatment of single-segment lower lumbar vertebral disease. METHODS: A total of 78 patients with various lower lumbar single-segment vertebral diseases were treated in our department from January 2008 to December 2011. There were 21 males and 57 females, with an average age of 52.5 years, including lumbar disc degeneration (28 cases), local recurrence of lumbar disc herniation (9 cases), huge lumbar disc herniation (11 cases), lumbar disc herniation with spinal stenosis (13 cases), and lumbar degenerative spondylolisthesis (degree I) (17 cases). All patients were treated by unilateral pedicle screws in the median incision combined with contralateral gunsight-guided translaminar facet screw fixation and interbody fusion. RESULTS: No wound infection or skin necrosis around incision was observed after operation. No leakage of cerebrospinal fluid or nerve injury occurred during and after operation. Excepting 4 cases, 74 cases were followed up for 18-60 months, averaged 33.5 months. All but one patient (98.6%) received interbody fusion. The intervertebral height of the indexed level was well restored and maintained. At final follow-up. During follow-up, there was no screw loosening or pedicle fracture observed. No apparent degeneration of adjacent segments. The mean Japanese Orthopaedic Association (JOA) scores was increased significantly from 12.79 ± 2.12 preoperatively to 25.8 ± 2.87 at the final follow-up. CONCLUSION: Unilateral pedicle screw fixation combined with contralateral translaminar facet screw fixation and interbody fusion showed good mid-term outcomes in the treatment of single-segment lower lumbar vertebral disease, and can be used as an optimal choice for fixation and fusion of some single-segment lower lumbar vertebral diseases.

Combination of percutaneous unilateral translaminar facet screw fixation and interbody fusion for treatment of lower lumbar vertebra diseases: a follow-up study.

OBJECTIVE: To investigate the feasibility and efficacy of sight-guided percutaneous unilateral translaminar facet screw fixation (TLFSF) with interbody fusion for treatment of lower lumbar vertebra diseases. METHODS: Twenty-nine adult patients with lower lumbar disease underwent sight-guided percutaneous unilateral TLFSF combined with microsurgical spinal decompression, discectomy, and interbody fusion from June 2007 to June 2008. All 29 patients had low back pain caused by lumbar disc degeneration (20 cases), in situ recurrent lumbar disc herniation (2), primary diskitis (1), lumbar disc herniation with spinal stenosis (3), and first-degree lumbar degenerative spondylolisthesis (3). Twenty-three cases had lesions at L4,5 ; three at L5 S1 , one at L3,4 , L4,5 , and two at L4,5 , L5 S1 . RESULTS: No patient experienced significant postoperative complications. The mean incision length was 4.48 ± 0.55 cm; operative time 1.34 ± 0.22 h; intraoperative blood loss 280 ± 175 mL; and postoperative drainage volume 165 ± 85 mL. Screw position results: type I, 21 cases (23 segments); type II, 7 cases (8 segments); and type III, 1 case (1 segment). Twenty-eight patients were followed up for 24-60 months (average, 47.5 months). Interbody fusion rate was 93.5%). Postoperative intervertebral height recovered significantly; however, loss of intervertebral height occurred during follow-up. CONCLUSION: Sight-guided percutaneous unilateral TLFSF with interbody fusion for treatment of lower lumbar disease is simple and minimally invasive, with good screw accuracy and security, high fusion rate, and good efficacy. However, specific surgical indications must be strictly followed.

[Injury characteristics and surgical treatment of lower lumbar vertebral burst fractures].

OBJECTIVE: To investigate the feasibility of utilizing self-designed score system for lower lumbar vertebral burst fractures to select surgical approach. METHODS: From January 2006 to December 2011, the clinical data of 56 patients with lower lumbar vertebra burst fractures who underwent surgical treatment were retrospectively analyzed. There were 42 males and 14 females with an average age of 43.1 years old (ranged, 19 to 65). Causes of injury included falling down (40 cases), traffic accidents (12 cases), and crashing injury by heavy objects(4 cases). Injury site was L3 in 37 cases, L4 in 16 cases, and L5 in 3 cases. According to the AO classification, 17 cases were type A3.1, 14 cases were type A3.2, 25 cases were type A3.3. According to Frankel grade of nerous function, 2 cases were grade B, 5 cases were grade C, 9 cases were grade D and 40 cases were grade E. Surgical methods and approaches were chosen based on the comprehensive evaluation of AO classification, condition of posterior column injury and spinal canal encroachment. Surgical methods and approaches included trans-vertebra fixation (15 cases), intra-vertebra pedicle screw fixation (21 cases), combination of anterior and posterior approaches (11 cases), one-stage posterior approaches (9 cases). Cobb angles, restorations of the affected vertebral anterior border height, and conditions of spinal canal encroachment were compared before and after surgery. Conditions of bone graft fusion and internal fixation (if bending, loosening or breakage existed) were observed. Spinal cord functions were assessed according to Frankel grade. Localized pain and working status of patients were also assessed at the last follow-up. RESULTS: No incision infection was found and no spinal nerve symptoms improved in all of 56 patients. All patients were followed up for 12 to 60 months with a mean of 28.5 months, without internal fixation loosening or breakage. There was significant differences in Cobb angle, vertebral anterior border height and recovery of spinal canal encroachment between preoperative and postoperative instantly (P < 0.05), however, there was no significant difference between postoperative instantly and final follow-up (P > 0.05). Thirteen cases obtained fusion by trans-vertebra fixation, 20 cases obtained fusion by intra-vertebra fixation, and 20 cases were treated by the combination of anterior and posterior approaches or one-stage posterior approaches all of patients obtained fusion. Spinal nervous function recovered I to II grade, 1 case was grade C, 3 cases were grade D, 52 cases were grade E. Localized pain was assessed as P1 in 52 cases, P2 in 3 cases, and P3 in 1 case. Working status was classified into W1 in 12 cases, W2 in 39 cases, and W3 in 5 cases. CONCLUSION: The lower lumbar vertebra and thoracolumbar junction exhibit different injury characteristics due to variations in anatomy and biomechanics. A comprehensive score of the AO classification, posterior column injury and degree of spinal canal encroachment will guide the selection of surgical method and approach for the treatment of lower lumbar vertebra burst fractures.