The cortical somatosensory evoked potentials (CSEP) is one of the electrophysiological examinations to determine the sensory pathways in the spinal cord, which can be used to quantify spinal cord function more accurately and is useful for the determination of spinal cord disorders, intraoperative monitoring, and prognostic assessment [1, 2, 3]. The authors selected 101 cases of spinal cord disorders in our department between August 1999 and April 2001 and performed preoperative examination, intraoperative monitoring, and postoperative spinal cord function recovery to analyze the correlation between the factors in order to further explore the value of CSEP in the field of orthopedics. 1. Clinical data and methods (1) General data In this group of 101 cases, there were 57 males and 44 females, aged 7-72 years old, with an average of 42 years old. The lesions were located in the cervical segment in 54 cases and in the thoracic and thoracolumbar segments in 47 cases, including spinal cord type cervical spondylosis, cervical disc herniation, cervical spine trauma, thoracic spinal stenosis, thoracic spinal canal occupied lesions (epidural and extramedullary subdural), thoracolumbar segment fracture dislocation, scoliosis deformity, and spinal tuberculosis and tumor (see Table 1). The procedures include anterior cervical decompression and implant fusion, posterior cervical single-opening vertebroplasty, thoracic laminectomy laminectomy decompression, thoracic spinal canal exploration and removal, thoracolumbar fracture repositioning and internal fixation, deformity correction, and spinal tumor resection. (2) Preoperative examination of CSEP The purpose of preoperative examination is to understand whether there is injury to the spinal cord and to provide a basis for preoperative diagnosis, and to record waveforms for reference and comparison after anesthesia and during surgery. The NECOLET evoked potential meter, version 1.7, was used. For the detection of cervical spinal cord, the ulnar or median nerve was stimulated with surface electrodes, and the recording lead was C3 or C4-FPZ; for the detection of thoracic and thoracolumbar spinal cord, the posterior tibial nerve was stimulated, and the recording lead was CZ-FPZ. The stimulation intensity was about 25-35 MA, the stimulation frequency was 3 HZ, and the number of superimposed waves was about 30-200 times. The latency, wave amplitude and differentiation of each waveform were observed. (3) Intraoperative CSEP monitoring Continuous epidural anesthesia or general anesthesia with amiflurane, laughing gas and fentanyl is generally used. During anesthesia, monitors are applied to continuously monitor blood pressure, heart rate, electrocardiogram, and oxygen saturation, and autologous blood transfusion is mostly used for complex surgery. The first measurement is performed before skin incision to record the performance of CSEP after anesthesia, and dynamic monitoring is performed intraoperatively before operations that may injure the spinal cord; the effects of other factors such as position change and external environment on CSEP are also observed so that valuable CSEP changes can be detected in a timely manner. (4) Intraoperative monitoring criteria Criterion Ⅰ: It is a generally accepted criterion, i.e., 10% prolongation of latency and/or 50% decrease in wave amplitude compared with post-anesthesia as the threshold of “alarm”. Criterion Ⅱ: The threshold value for intraoperative monitoring of clinical CSEP recommended by Shen was applied, i.e., intraoperative wave amplitude decreased by no more than 50% and latency prolonged by no more than 70% in patients of grade D and E, and intraoperative wave amplitude decreased by no more than 40% and latency prolonged by no more than 50% in patients of grade B and C [4]. In addition, spinal cord injury (reversible or non-reversible) can be diagnosed if one of the following conditions is met: (i) confirmed by intraoperative arousal test; (ii) postoperative clinical examination reveals increased functional impairment of the spinal cord. (5) Postoperative evaluation This study focused on assessing the clinical outcome of patients in the short term after surgery. During the patient’s hospitalization or after discharge (two weeks to six months), the remission was followed up, based mainly on the patient’s subjective symptoms and clinical signs examination. The treatment was judged to be effective when the subjective symptoms were relieved and there were corresponding changes in physical signs on examination. 2, results According to the test results, the waveform performance after spinal cord compression or injury can be divided into the following four types: (1) type No wave amplitude is elicited, and it resembles a straight line. The three cases in this group were complete spinal cord injuries, all of which were Frankel A patients, and intraoperative monitoring was of little significance; the spinal cord function did not recover significantly after surgery, and the purpose of surgery was mostly to stabilize the spine rather than to rescue the spinal cord. （Type (2) Obviously abnormal, visible response, but the wave sequence cannot be identified, or the wave positive and negative in an irregular form, the latency can not be calculated or inaccurate calculation. There were 21 cases in this group, which were partial spinal cord injuries and were patients with Frankel grade B (4 cases), C (15 cases) or D (2 cases). Intraoperative monitoring was not ideal; 11 patients had recovery of sensory muscle strength after surgery, and most of them had acute spinal cord injury, and the treatment efficiency was 52.4%. (3) type CSEP abnormalities, positive and negative waves appear, measurable, good differentiation of each wave, prolonged latency and decreased wave amplitude. In this group, there were 59 cases with mild to moderate spinal cord injury, which were patients with Frankel grade B (1 case), Frankel grade C (14 cases) or grade D (44 cases). Intraoperative monitoring is of great significance to detect abnormal changes in time; 18 cases were evaluated for changes according to diagnostic criteria I, and 13 cases were evaluated for abnormal intraoperative performance according to diagnostic criteria II, and 9 cases were confirmed to have reversible spinal cord damage. Most of the patients’ symptoms were relieved after the operation, and 49 patients had significant recovery of sensory muscle strength, with an efficacy of 83.1%. (4) Type Normal waveform, peak wave slightly W-shaped, its latency and wave amplitude are in the normal range. In this group of 19 cases, the spinal cord was not significantly damaged and was Frankel grade E (4 cases) or D (15 cases), which is the early stage of the lesion. The purpose of intraoperative monitoring was clear to reduce complications, especially the occurrence of medically induced paraplegia; 6 cases had changes according to diagnostic criteria I, and 4 cases had abnormal intraoperative performance according to diagnostic criteria II, while 2 cases were confirmed to have reversible spinal cord damage. After surgery, 16 cases had different degrees of remission, and the efficacy was 84.2%. The data related to CSEP typing and Frankel classification were analyzed as shown in Table 2 and Figure 2: all patients with CSEP type I manifestation were Frankel A, type II were mainly B and C, type III were mainly C and D, and type IV were mainly D and E. All patients with CSEP type A manifested type I, most of type B manifested type II, type C manifested type II and III, type D manifested type III and IV, and type E manifested type IV. The correlation between CSEP typing and postoperative symptom relief rate is shown in Table 3, and the CSEP type I patients showed no significant improvement, type II had fair efficacy, and type III and IV had more positive efficacy. According to the comparison of the number of cases of intraoperative monitoring results and evaluation indexes of different diagnostic criteria, see Tables 4 and 5 respectively, we can see that the sensitivity and leakage rate are the same, and the specificity and misdiagnosis rate are different. In addition, the analysis of 11 cases of intraoperative CSEP changes (reversible damage to the spinal cord was confirmed), mainly resection of vertebral tumors (3 cases), removal of intracanal space-occupying lesions (2 cases), lateral bracing of scoliosis (2 cases), stretching of the spinal cord during “single opening” of the posterior cervical spine (2 cases), laminectomy for thoracic spinal stenosis (1 case), and placement of a thoracic CD plate hook (1 case). The percentages of each of the operations that could cause changes in CSEP are shown in Figure 3. Since 1977, when somatosensory evoked potentials were used for intraoperative monitoring, CSEP has been increasingly used in the field of spinal cord surgery.CSEP is conducted by sensory impulses through the posterior cord of the spinal cord, i.e., the thin bundle and the cuneiform bundle, and because the sensory area of the spinal cord is close to the anterior horn of the spinal cord and is surrounded by the arachnoid membrane as a whole, CSEP examination can detect spinal cord injury and its extent in a timely manner [5]. (1) Correlation between the degree of spinal cord injury, prognosis and CSEP examination results The degree of spinal cord injury can be quantified by Frankel’s classification, but it is often influenced by subjective factors. In this group of cases, the CSEP examination results are consistent with the spinal cord grading. Patients with type I manifestations are mostly spinal cord transection and complete paresis, and surgical decompression is not significant for spinal cord recovery, and the postoperative symptom relief is not satisfactory; patients with type II manifestations are mostly those with severe spinal cord damage and incomplete paresis, and surgery can create certain conditions for spinal cord recovery, but the efficacy is uncertain; patients with type III manifestations have mild spinal cord damage, and surgical decompression is significant for spinal cord recovery. CSEP is an important tool to diagnose spinal cord injury and determine the evolution of reversible spinal cord injury, as it can provide a more accurate functional diagnosis and quantitative analysis of the spinal cord. The present group of cases showed that CSEP findings correlated well with the degree of spinal cord injury, but also with the rate of postoperative symptom relief, which to some extent helps in the choice of treatment. In addition, CSEP has a central amplification effect, is very sensitive, and responds to changes in the condition 3-4 weeks earlier than clinical signs, so the prognosis can be judged according to the degree of shortening of the latency period [6]. (2) CSEP monitoring in spinal cord surgery Early detection of reversible spinal cord injury, determination of the degree of injury, and effective prevention of medically induced paraplegia are important objectives of CSEP monitoring in spinal cord surgery. It can identify acute injuries and their sites in the nerve conduction pathways, and correct the causative factors in a timely manner, thus greatly improving the safety of spinal cord surgery; at the same time, it also makes it possible to operate in high-risk or even forbidden areas. In this group of cases, the surgical operations that are prone to spinal cord injury are, in order of priority: resection of vertebral tumors or occupying lesions in the spinal canal, orthopedic scoliosis with concave lateral bracing, posterior cervical “single-opening” vertebroplasty, removal of the thoracic spinal stenosis lamina, and placement of the lamina hook under the thoracic lamina. In the author’s opinion, CSEP has the most obvious value in spinal cord tumor resection and scoliosis orthopedic surgery. During scoliosis orthopedics, especially concave bracing, the spinal cord and blood vessels are susceptible to stretching changes, and excessive stretching can lead to local ischemia in the spinal cord, and CSEP is sensitive to the ischemic changes in the spinal cord, immediately showing prolonged latency, decreased amplitude, and even loss of waveform [7]. During tumor resection, CSEP can also help to identify the nerve tissue around or within the tumor, ensuring that the surgeon performs a more extensive and optimal operation. In addition, monitoring documentation is of potential forensic value to surgeons and anesthesia personnel, and procedures without monitoring may be considered to lack safety standards. (3) CSEP monitoring criteria Currently, CSEP monitoring is limited to observation of wave amplitude and latency changes. A 10% increase in latency and/or a 50% or greater decrease in wave amplitude indicates impairment of spinal cord function and is currently the generally accepted “gold standard”. In fact, it is difficult to establish a single standard due to the severity of the disease, the location of the lesion, and the surgical operation. In the present study, the sensitivity of different monitoring criteria was 100%, while the specificity was significantly different, and criterion II was more practical. In the author’s opinion, monitoring criteria are flexible within a certain range, varying with disease, site and operation, and it is a challenging task to detect changes in waveforms and reveal their causes. While it is important to detect valuable changes in a timely manner, too much information may interfere with the operator’s operation; and wrong information may lead to irreversible and serious consequences, and no information is better than wrong information. When the CSEP waveform changes to an “alarm value,” the monitor should watch for every small change; it is also helpful to be familiar with each important step in the procedure and to communicate with the surgeon in simple, intuitive terms. In addition, they should be alert to the rare possibility that the motor bundle is damaged while the sensory bundle is intact, and the patient may have severe motor nerve damage without CSEP changes after surgery [8]; at the same time, they should pay attention to the establishment of the baseline wave before surgery and after anesthesia, and correlation analysis between the potential changes and the baseline wave during surgery and the clinical performance of the patient can reveal valuable information more accurately. (4) Relationship between CSEP and other evoked potentials In addition to CSEP, other monitoring techniques developed in recent years include subcortical somatosensory evoked potentials (sub-CSEP), spinal somatosensory evoked potentials (SSEP) and motor evoked potentials (MEP). According to the literature [9], multipoint recording of somatosensory evoked potentials or the combined application of multiple modalities of SEP and MEP monitoring has good resistance to interference and not only responds better to the functional integrity of the spinal cord, but also has been used to respond to the depth of anesthesia, circulatory status, etc. CSEP records the electrophysiological activity generated by peripheral stimuli transmitted to the cerebral cortex via level 3-4 synapses, which is relatively less resistant to interference The CSEP is susceptible to the effects of anesthetics, temperature, blood pressure, partial pressure of carbon dioxide, as well as electric drills and electric knives; however, with the development of hardware and equipment, the sensitivity and interference resistance of CSEP monitoring have been significantly enhanced, and currently, as long as a ground wire is connected intraoperatively, other factors have less influence on the CSEP, except for the need to suspend monitoring when using an electric knife. In addition, CSEP does not require exposure of the dura mater or stimulation of the cerebral cortex, which makes it easier and safer to use for preoperative examination and postoperative evaluation, showing its unique appeal in the diagnosis and treatment of spinal cord disorders.