Volume 42 · Issue 04 · 2026

hypertensive intracerebral hemorrhage (HICH) refers to a type of cerebral hemorrhage that suddenly occurs in the brain parenchyma (basal ganglia, thalamus, ventricles, cerebellum and brainstem), primary (excluding vascular structural abnormalities, coagulation dysfunction, hematological diseases, systemic diseases and neoplastic diseases), non-traumatic, and has a clear history of hypertension^[1]。 HICH is one of the most serious complications of hypertension, accounting for 10% to 30% of stroke, and the 6-month mortality rate is as high as 30% to 50%^[2]It is one of the common emergencies in neurology and surgery. The treatment of HICH often requires the close cooperation of emergency department, neurosurgery, neurology, anesthesiology, critical care medicine and rehabilitation department; Critically ill patients often require supportive treatment from the Department of Critical Care Medicine.

Peripheral facial nerve paralysis (facial paralysis) is a common clinical disease, which is mainly manifested by facial weakness symptoms such as weakness of eyebrow lift, incomplete eye closure, shallow or even disappearance of nasolabial folds, bulging gills and air leakage, and drooping mouth corners^[1,2]。 Most patients can obtain a good prognosis, but some patients can't fully recover from facial nerve paralysis due to irregular, untimely treatment, and serious disease, and often develop symptoms such as facial muscle contracture, involuntary twitching, joint movement, small eye fission, difficulty in opening eyes, facial stiffness and tightness 3~6 months after the onset of the disease, and the above symptoms gradually aggravate with the prolongation of time, and the clinical diagnosis is a sequela of spastic facial paralysis^[3]。 Among them, joint movement is the main cause of facial discomfort^[4]。 Synchronous movement refers to the abnormal involuntary movement of another facial muscle caused by the voluntary movement of one facial muscle, which is mostly distributed around the eyes, cheeks and mouth. Among them, mouth-eye, eye-mouth synchronous movement is the most common type, and others include eye-chin, eye-nose, chin-mouth, chin-eye synchronous movement and other types^[5]。 Mild joint exercise will cause patients to eat, drink, brush teeth, gargle and other daily activities disorders, and severe joint exercise will also lead to abnormal facial appearance, so patients are unwilling to participate in any social activities, and even have social fear and severe depression^[6]。

To construct a prediction model for poor long-term neurological outcome in patients with traumatic brain injury (TBI) and severe polytrauma undergoing craniocerebral surgery and to verify its predictive performance.

The clinical data of 717 patients with TBI and severe polytrauma who received surgical treatment at the Neurosurgery Department of Changde Hospital Affiliated to Xiangya Medical College of Central South University (Changde First People′s Hospital) from May 2019 to March 2025 and completed 1-year postoperative follow-up were retrospectively analyzed. A simple randomization method was used to divide the patients into a training set (501 cases) and a validation set (216 cases) at a ratio of 7 ∶3. The 1-year neurological functional outcome after craniocerebral surgery was used as the research end point. Based on the Glasgow outcome scale-extended(GOS-E) score at 1 year after surgery, the patients in the training set were divided into the good outcome group (GOS-E: 5 to 8 points, 191 cases) and the poor outcome group (GOS-E: 1 to 4 points, 310 cases). The patients′ baseline information on admission, treatment and postoperative complications were collected. In the training set, univariate analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were used to screen risk factors, and they were included in multivariate logistic regression analysis to determine the risk factors for poor long-term neurological function prognosis in patients with TBI and severe polytrauma after craniocerebral surgery, and a regression equation was constructed to draw a visual nomogram model. In the training set and validation set, the predictive performance of the model was evaluated by receiver operating characteristic curve and area under the curve (AUC), calibration curve and Brier score, and clinical decision curve analysis (DCA).

Risk factor screening results showed that age, the time between trauma and craniocerebral surgery, pupillary photoreactivity on admission, the abbreviated injury scale of the head and neck, intracerebral hematoma, blood transfusion volume during craniocerebral surgery, traumatic cerebral infarction, activated partial thromboplastin time on admission, and blood sugar on admission were all independent risk factors for poor long-term neurological prognosis after craniocerebral surgery in patients with TBI and severe polytrauma (all P<0.05). The constructed nomogram prediction model had an AUC of 0.77 (95%CI: 0.73-0.81) in the training set, a Youden index of prediction probability of 0.58, a sensitivity of 75.5%, and a specificity of 82.0%; in the validation set, the AUC was 0.72 (95%CI: 0.65-0.79), the Youden index of the prediction probability was 0.60, the sensitivity was 67.2%, and the specificity was 75.9%. Calibration curve analysis showed that the Brier score of the training set was 0.19 (95%CI: 0.17-0.20), and the Brier score of the validation set was 0.20 (95%CI: 0.18-0.23). The DCA results showed that the threshold probabilities of the training set and the validation set were both in the range of 0 to 98%, and the net return rate was >0, suggesting that the nomogram prediction model had good clinical application value.

The nomogram model constructed based on the risk factors for poor neurological prognosis after craniocerebral surgery in patients with TBI and severe polytrauma has certain predictive ability and could help identify patients with a higher risk of poor long-term prognosis.

To develop a predictive model for the risk of intracranial aneurysm (IA) rupture based on clinical and imaging data and to evaluate its predictive performance.

This retrospective study included 645 patients with IA admitted to the Department of Critical Care Medicine, First Affiliated Hospital of Xinjiang Medical University from January 2017 to December 2023. Baseline information, clinical and laboratory data, and imaging data were collected. Patients were randomly divided into a training set (n=454) and a validation set (n=191) at a 7 ∶3 ratio. Univariate and multivariate logistic regression analyses were performed in the training set to identify risk factors for IA rupture. A nomogram model was constructed using R software. The predictive performance of the model was evaluated using the receiver operating characteristic (ROC) curve and the area under the curve (AUC), calibration curve, and decision curve analysis (DCA).

Among the 645 patients, 315 had ruptured IAs, with 223 in the training set and 92 in the validation set. A total of 8 factors were identified as independently associated with IA rupture. Elevated diastolic blood pressure, blood glucose, triglyceride-glucose index and high-density lipoprotein cholesterol were risk factors for IA rupture, whereas increased hematocrit, higher body mass index, regular aneurysm morphology and location in internal carotid artery were protective factors. The model achieved an AUC, sensitivity and specificity of 0.84, 0.88 and 0.65 in the training set and 0.86, 0.88 and 0.75 in the validation set respectively. The calibration curves and DCA results indicated good calibration and clinical utility for the prediction model in both sets.

The nomogram model constructed by integrating clinical and imaging features demonstrates high efficacy in predicting the rupture of IA.

To investigate the safety and efficacy of one-stage ventriculoperitoneal shunt (VPS) combined with cranioplasty in the treatment of skull defect complicated with hydrocephalus.

A retrospective analysis was performed on the clinical data of patients with skull defect and hydrocephalus who underwent one-stage VPS combined with cranioplasty in the Department of Neurosurgery, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences from January 2017 to December 2024. At 3 months after surgery, postoperative recovery were evaluated based on cranial imaging and Glasgow Coma Scale (GCS) score. The occurrence of complications were observed.

A total of 25 patients were enrolled, with a preoperative GCS score of 10.40±3.24. All one-stage operations were successfully completed. Postoperative cranial CT showed that the cranioplasty patches were all in good position and intact, the morphology of brain tissue was improved compared with preoperative status; the positions of the drainage tube and valve were all satisfactory, and ventricular dilatation was significantly relieved. The GCS score at 3 months postoperatively was 10.20±3.22, with no statistically significant difference compared with the preoperative level (paired t=0.82, P=0.416). During perioperative period, fever occurred in 6 patients (24%), including 5 cases of pulmonary infection and 1 case of urinary tract infection, all of which were resolved after treatment. Subcutaneous effusion at the surgical site occurred in 3 patients (12%), all of whom were cured after puncture aspiration and posture adjustment. No severe complications such as intracranial hemorrhage, subdural effusion, intracranial infection or cerebrospinal fluid leakage were observed.

One-stage VPS combined with cranioplasty is safe and effective in the treatment of skull defect complicated with hydrocephalus.

To evaluate the clinical efficacy and safety of navigation-assisted neuroendoscopic surgery based on three-dimensional (3D) reconstruction in patients with acute cerebral hemorrhage.

A retrospective analysis was conducted on the clinical data of patients with acute intracerebral hemorrhage admitted to the Department of Neurosurgery, Second Affiliated Hospital of Zhengzhou University from January 2020 to December 2023. Among them, 75 patients who underwent navigation-assisted neuroendoscopic surgery based on 3D reconstruction were assigned to the 3D reconstruction group, and 75 patients who underwent conventional neuroendoscopic surgery were included as the control group using a propensity score matching strategy at a 1 ∶1 ratio based on similar clinical characteristics. The clinical efficacy and safety of the two groups were evaluated. Efficacy indicators included operation duration, intraoperative blood loss, hematoma clearance rate, postoperative hospital stay, as well as modified Rankin scale (mRS), National Institutes of Health Stroke Scale (NHISS), Activities of Daily Living (ADL) and Glasgow Coma Scale(GCS) scores. The safety indicator was the occurrence of complications in patients.

There were no statistically significant differences in preoperative GCS, ADL, NIHSS, and mRS scores between the two groups (all P>0.05). The preoperative preparation time in the 3D navigation group was longer than that in the control group (25.84±4.91 min vs. 12.49±2.95 min, P<0.001). Regarding surgical approaches, the 3D navigation group predominantly used function-preserving trajectories, while the control group predominantly used shortest trajectories, with statistically significant difference between the two approaches (P<0.001). The overall effective rate[96.0% (72/75) vs. 81.3%(61/75)] and hematoma clearance rate [(91.34±1.79)% vs. (77.82±1.11)%] in the 3D reconstruction group were significantly higher than those in the control group (both P<0.01). The operation duration (99.12±4.27 min vs. 110.68±4.65 min), intraoperative blood loss(38.60±1.05 ml vs. 44.31±1.30 ml), and hospital stay in the 3D reconstruction group (13.44±0.91 d vs. 15.70±1.08 d) were significantly lower than those in the control group (all P<0.01). At 1 and 3 months after surgery, the GCS scores of patients in the 3D reconstruction group were higher than those in the control group (both P<0.01), while there was no significant difference in GCS scores between the two groups at 6 months after surgery (P=0.348). One month after operation, the ADL scores of patients in the 3D reconstruction group significantly increased, while the NHISS and mRS scores significantly decreased, with statistical significance compared with the control group (all P<0.01). The overall complication rate in the 3D reconstruction group was 2.7% (2/75), which was significantly lower than that in the control group [13.3% (10/75)] (P=0.002).

Navigation-assisted neuroendoscopic surgery based on 3D reconstruction demonstrates good efficacy and a lower risk of complications in non-comatose patients with acute intracerebral hemorrhage, and it is worthy of clinical promotion and application.

To explore the predictive value of intracranial pressure variability for the prognosis of elderly patients with intracerebral hemorrhage after surgery.

A retrospective analysis was conducted on the clinical data of 320 elderly patients with acute intracerebral hemorrhage who underwent surgery at the Department of Neurosurgery, Affiliated Hospital of North China University of Science and Technology from May 2024 to January 2025. Based on the 30-day postoperative modified Rankin scale (mRS) scores, patients were divided into the good outcome group (mRS score<4; 151 cases) and the poor outcome group (mRS score≥4; 169 cases). The two groups were compared in terms of general information, preoperative vital signs, emergency cranial CT results, admission Glasgow coma scale (GCS) scores, emergency laboratory blood test indicators, and intracranial pressure (ICP)-related indicators (mean and IPV indicators, with IPV indicators including standard deviation, successive of variation [SV], and coefficient of variation of ICP). A univariate Logistic regression model was used to analyze the relationship between IPV indicators and prognosis; subsequently, indicators with P<0.05 in the group comparison were included in a multivariate Logistic regression model for adjustment to further analyze whether IPV indicators were independent factors affecting prognosis. Receiver operating characteristic curves were used to evaluate the predictive efficacy of each IPV indicator for poor postoperative prognosis in elderly patients, and the Delong test was employed to compare the differences in predictive performance among various predictive factors.

Compared with the good outcome group, patients in the poor outcome group were older, had a higher proportion of hemorrhages located in the infratentorial region, and exhibited higher levels of total protein, low-density lipoprotein, lactate dehydrogenase, homocysteine, and ICP-related indicators. Conversely, the proportions of hemorrhage breaking into the ventricles and midline shift, as well as admission GCS scores, albumin, and high-density lipoprotein levels, were lower. All these differences were statistically significant (all P>0.05). Univariate Logistic regression analysis showed that increased SV, standard deviation, and coefficient of variation of ICP were all risk factors for poor prognosis (all P<0.001). Multivariate logistic regression analysis indicated that after adjusting for differences between the two groups, the SV(OR=1.44, 95%CI: 1.32-1.57), standard deviation (OR=3.40, 95%CI: 2.42-4.78), coefficient of variation (OR=1.18, 95%CI: 1.13-1.23) of ICP were all independent risk factors for poor prognosis (all P<0.001). The areas under the curve (95%CI) for predicting poor postoperative outcomes in elderly patients with acute intracerebral hemorrhage were 0.90 (0.86-0.93) for the SV, 0.80 (0.75-0.85) for the standard deviation, and 0.81 (0.76-0.85) for the coefficient of variation. The Delong test results showed that the predictive efficacy of the SV of ICP was higher than that of the standard deviation and coefficient of variation of ICP (both P<0.05).

Elevated IPV indicators could effectively predict poor postoperative prognosis in elderly patients with acute cerebral hemorrhage, with the best predictive effect being the SV of ICP.

The clinical data of 9 pediatric patients with epidural hematoma admitted to the Department of Neurosurgery of the First Affiliated Hospital of Zhengzhou University from September 2023 to September 2024 were retrospectively analyzed. CT or MRI examination of the head was performed before operation, and the surgical incision was designed according to the location, shape and bleeding volume of the hematoma. During the operation, the hematoma was removed by a "cross" shaped bone hole incision under neuroendoscope. Hematoma volume decreased on postoperative day 1 compared to preoperative [4.44 (1.11, 5.76) ml and 30.57 (22.86, 40.73) ml, respectively,Z＝－2.666，P＝0.008]。 Glasgow Coma Scale scores increased on postoperative day 1 compared with preoperative scores (15.0 (14.0, 15.0) points and 12.0 (9.5, 13.5) points,Z＝－2.536，P＝0.011]。 Nine children were followed up for (7.6±5.4) months (3 to 12 months). By the last follow-up, the Glasgow prognostic grade of all nine children was V. CT or MRI revealed no residual hematoma at 6 months after surgery. For children's epidural hematoma, the use of neuroendoscopic "cross" bone hole incision surgery to treat it is less invasive and clinical efficacy is definite.

To investigate the characteristics of postoperative fever after acoustic neuroma resection via a suboccipital retrosigmoid approach and its relationship with clinical factors, blood routine, blood biochemical parameters, and cerebrospinal fluid (CSF) test results.

A retrospective consecutive cohort of 195 patients who underwent vestibular schwannoma resection via the retrosigmoid suboccipital approach at the Neurosurgery Center, Beijing Tiantan Hospital, Capital Medical University between January and June 2024 was analyzed. Fever was defined as the body temperature ≥37.2 ℃, and marked fever as ≥38.5 ℃. Patients were accordingly classified into a marked-fever group and a non-marked-fever group. Differences between groups in sex, age, time to fever onset (≥37.2 ℃), operative duration, tumor laterality, and tumor volume were compared. Variables with P<0.20 in univariate analyses were entered into a multivariable logistic regression model. Univariate linear regression was used to assess associations between relevant variables and the duration of marked fever. Dynamic changes in blood and CSF parameters were analyzed and compared between the two groups.

Among the 195 patients, 182 (93.3%) developed postoperative fever, with a median onset of 1 day (range: 0-11 days) after surgery. Marked fever occurred in 25 patients (12.8%), with a median onset of 6 days postoperatively (range: 0-15 days), and a median duration of 5 days (range: 1-19 days) until recovery to normothermia. Compared with the non-marked-fever group, patients with marked fever were younger, had a later onset of postoperative fever, and had longer operative duration (all P<0.05). Multivariable logistic regression analysis demonstrated that time to fever onset (OR=1.349, 95%CI: 1.027~1.871, P=0.034) and tumor volume (OR=1.049, 95%CI: 1.007-1.095, P=0.022) were independently associated with the development of marked postoperative fever. Operative duration was positively correlated with the duration of marked fever (β=1.031, 95%CI: 0.018-2.044, P=0.046). In the marked-fever group, blood inflammatory markers including total white blood cell count, neutrophil count, neutrophil percentage and C-reactive protein returned to normal levels by postoperative day 6, followed by a marked secondary increase. In contrast, these parameters normalized by postoperative days 4-5 in the non-marked fever group, with only minor subsequent fluctuations. At postoperative day 6, the incidence of hypokalemia and hyponatremia as well as CSF white blood cell counts showed a significant upward trend in the marked fever group. At postoperative day 4, the proportions of mononuclear and polymorphonuclear cells in CSF differed most significantly between the two groups.

Postoperative fever is common after vestibular schwannoma resection. Patients with a later onset of fever and larger tumor volume are more likely to develop marked fever, while longer operative duration is associated with a prolonged duration of marked fever. Patients with marked fever exhibit greater fluctuations in blood inflammatory markers and CSF white blood cell counts.

To investigate the pathophysiological mechanism of trigeminal neuralgia (TN) by analyzing changes in brain function in TN patients using resting-state functional magnetic resonance imaging (rs-fMRI).

Clinical data and rs-fMRI images of 93 TN patients (TN group) who underwent surgical treatment at Department of Neurosurgery, Peking University People′s Hospital from July 2024 to April 2025 were prospectively analyzed. Thirty-two healthy controls (HC group) matched for age, sex and education level were included as controls. Rs-fMRI data were processed using Matlab, SPM12 and RESTplus software to calculate the amplitude of low-frequency fluctuation (ALFF), fractional ALFF (fALFF), regional homogeneity (ReHo) and functional connectivity. Two-sample t-tests were used to identify brain regions with differences between the TN and HC groups. Pearson correlation analysis was then performed to explore the relationships of these regions with disease duration, affected side, involved region, presence of persistent pain, Visual Analog Scale (VAS) score, Self-Rating Depression Scale score, Self-Rating Anxiety Scale score, Pittsburgh Sleep Quality Index score and surgical success.

There were no statistically significant differences between the TN group and HC group in terms of age, gender and education level (all P>0.05). Compared with the HC group, TN patients showed abnormal functional activity in multiple brain regions, including the right fusiform gyrus, bilateral medial superior frontal gyri, left superior occipital gyrus, right thalamus, bilateral posterior cingulate gyri and left caudate nucleus. The ReHo value of the right fusiform gyrus was negatively correlated with the VAS score (r=-0.312, P=0.002), while the ReHo value of the right thalamus was positively correlated with the VAS score (r=0.216, P=0.011). No significant correlations were found between functional indices in the other differential brain regions and clinical variables (all P>0.05).

Multiple brain regions in TN patients, including the thalamus and cerebellum, showed abnormalities in neural spontaneous activity, coordination and functional connectivity during resting state, which may be associated with the occurrence and development of pain perception and central pain sensitization.

To explore the application techniques and clinical efficacy of retrograde negative-pressure aspiration technique in the microsurgical treatment of large (maximum diameter 1.5-2.5 cm) or giant (maximum diameter >2.5 cm) paraclinoid internal carotid artery aneurysms.

A retrospective analysis was conducted on the clinical data of 21 patients with large or giant paraclinoid internal carotid artery aneurysms who underwent microsurgical treatment assisted by retrograde negative-pressure aspiration technique in the Cerebrovascular Disease Treatment Center, Brain Hospital Affiliated to Nanjing Medical University from July 2019 to July 2024. Through the cervical incision, the common carotid artery, external carotid artery, and proximal internal carotid artery were exposed and controlled, and a retrograde negative-pressure aspiration channel was established via a 5F arterial sheath. Simultaneously, the anterior clinoid process was drilled out through the intracranial pterional Dolenc approach to fully expose and manage the aneurysm. Postoperatively and during follow-up, the Glasgow Outcome Scale (GOS) grade was used to assess neurological function, with GOS ≥Ⅳ grade defined as a favorable outcome. Cranial CT and MRI were performed to evaluate the presence of new intracranial hemorrhage or cerebral infarction. Cranial CTA or DSA was used to assess the status of aneurysm clipping and the presence of residual or recurrent aneurysm.

A total of 21 aneurysms were treated in 21 patients, including 13 large and 8 giant paraclinoid internal carotid artery aneurysms. Complete clipping was achieved in all cases, and no residual aneurysm was found on postoperative CTA or DSA. A total 17 patients showed improvement in neurological function. New postoperative neurological deficits occurred in 2 patients, including 1 case of acute basal ganglia infarction and 1 case of delayed internal carotid artery occlusion that underwent decompressive craniectomy. No deaths occurred. At discharge, 19 patients had a GOS grade of Ⅴ, 1 had Ⅳ, and 1 had Ⅱ. The mean follow-up duration was 24.8±2.5 months (range: 5-40 months). During follow-up, there were no new intracranial hemorrhages, or cerebral infarctions. No residual or recurrent aneurysm was detected on CTA or DSA in any patient, indicating curative clipping. At the last follow-up, 19 patients had a GOS grade of Ⅴ, 1 had Ⅳ, and 1 had Ⅱ, with a favorable outcome rate of 95.2% (20/21).

The application of retrograde negative-pressure aspiration technique in the microsurgical treatment of large or giant paraclinoid internal carotid artery aneurysms is safe and effective, achieving a high rate of complete clipping and favorable functional outcomes. This technique provides a reliable therapeutic option for the surgical management of complex aneurysms.

To construct a mouse model for cerebral hemorrhage (ICH)-induced hemiplegia treated by deep brain stimulation (DBS) of the cerebellar dentate nucleus and to evaluate the effect of DBS on the motor function of the mice.

The right striatum was injected with type Ⅳ collagenase to induce intracerebral hemorrhage, thereby preparing a model of ICH-induced hemiplegia in mice. Based on whether collagenase was injected, the mice were divided into the ICH group and the sham operation group. On the basis of the ICH-induced hemiplegia mouse model, electrode implantation into the left cerebellar dentate nucleus electrode was further performed to prepare an ICH+ DBS hemiplegia mouse model; and current stimulation (intensity: 30 μA, pulse width: 60 μs, frequency: 30 Hz) was administered for 2 weeks; and the mice were divided into the ICH+ DBS (ON) group and the ICH+ DBS (OFF) group based on whether current stimulation was performed. The limb motor function of the mice was evaluated using the cylinder test, the corner test and the open field test.

The ICH (n=8) and sham operation mouse models (n=8) were successfully established, with a modeling success rate of 80%(16/20) for the ICH mouse model. The analysis of mouse motor function showed that there was a statistically significant difference in the cylinder test hemiplegia score between the ICH group and the sham operation group (F_groups=78.87, F_time=6.56, F_interaction=5.74; all P<0.001); at 3, 7, 14, 21, and 28 days after surgery, pencentage of affected timb use in cylinder test of the ICH group was lower than that of the sham operation group (all P<0.05). The ICH+ DBS mouse model was successfully established, with an electrode implantation success rate of 69.05%(20/29). At 21 days after surgery, compared with the ICH+ DBS (OFF) group (n=10), the cylinder test hemiplegia score of the ICH+ DBS (ON) group (n=10) increased [(31.25±12.38)% vs. (17.12±14.05)%], the asymmetry index of the turn angle test decreased (5.74±2.87 vs. 14.59±9.24); the movement distance in the open field test increased (76.73±13.54 m vs. 62.61±9.81 m), the maximum speed increased (0.382±0.048 m/s vs. 0.336±0.050 m/s), and all differences were statistically significant (all P<0.05).

The model of mouse brain hemorrhage combined with DBS implantation in the cerebellar dentate nucleus can be successfully established with a high success rate and stable phenotype. DBS in the cerebellar dentate nucleus could improve the upper limb motor function of mice with ICH.

To systematically evaluate the impact of post-stroke delirium (PSD) on mortality and length of hospital stay in acute stroke patients.

Computerized searches were conducted in PubMed, Embase, Psychological Science Database, Medline Database, China Journal Full-text Database and Wanfang Database for clinical studies on the impact of PSD on stroke patient mortality and length of hospital stay. The search period covered the time from the establishment of the databases to February 2025. The literatures were screened according to the inclusion and exclusion criteria, and data extraction and quality evaluation were performed. Meta-analysis was conducted using R language version 4.5.2 and heterogeneity among the studies was evaluated by the I² value. If I² was >50%, the random effects model was used for analysis; if I² was ≤50%, the fixed effects model was used employed for analysis.

A total of 25 studies were included, including 23 cohort studies and 2 case-control studies; a total of 10 068 patients were involved, among which 2 413 were in the delirium group and 7 655 were in the non-delirium group. The meta-analysis of the impact on patient mortality rates showed that 4 studies fully adjusted for confounding factors. Compared with the non-delirium group, PSD was significantly associated with an increase in patient mortality rates (OR=3.47, 95%CI: 2.35-5.12, P<0.001), with a median follow-up time of 7.5 months (3-60 months), and the heterogeneity of the results was extremely low (I²=0%); the combined analysis of the 25 included studies also showed that PSD was significantly associated with an increase in patient mortality rates (OR=3.83, 95%CI: 2.86-5.14, P<0.001), with a median follow-up time of 12.7 months (0-60 months), and the heterogeneity of the results was relatively high (I²=66.6%); the meta-analysis of the impact on patient hospital stay showed that 1 study fully corrected for confounding factors, and the result was that PSD was significantly associated with an increase in hospital stay (RR=1.63, 95%CI: 1.11-2.39, P<0.001); the data of the 17 studies were combined for analysis, and the result showed that PSD was significantly longer than that of patients without delirium in hospital stay (SMD=0.90, 95%CI: 0.37-1.43, P<0.001), and the heterogeneity of the results was relatively high (I²=98.5%). Subgroup analysis by stroke type showed that in patients with ischemic stroke, PSD could significantly increase the risk of patient death (OR=2.94, 95%CI: 1.30-6.66, P<0.001), but there was no statistical significance in the correlation with hospital stay (SMD=0.67, 95%CI: -0.11-1.02, P=0.090); in patients with intracerebral hemorrhage, PSD significantly increased the risk of patient death (OR=29.68, 95%CI: 7.20-122.34), and also significantly prolonged the hospital stay of patients (SMD=0.88, 95%CI: 0.70-1.05), and the correlations were statistically significant (all P<0.001).

PSD increases the risk of mortality for patients and is not affected by key confounding factors. There is low heterogeneity among the studies. Moreover, post-stroke delirium prolongs the hospital stay of stroke patients, and there is high heterogeneity among the studies.

Trousseau syndrome is an arteriovenous thromboembolic event triggered by a hypercoagulable state associated with a malignant tumor and is more common in patients with lung cancer, gastrointestinal cancer and gynecological advanced adenocarcinoma. Trousseau syndrome with acute and multiple massive cerebral infarction as the first manifestation is rare in clinic. This article reports a patient with multiple massive cerebral infarction as the first manifestation and finally diagnosed as lung adenocarcinoma complicated with Trousseau syndrome. During anticoagulation treatment, hemorrhage transformation occurred, and the treatment was abandoned due to poor general condition, and died 3 days after discharge.

Endovascular therapy (EVT), as the core treatment for acute ischemic stroke, has significantly improved the functional prognosis of patients. However, the mortality rate of patients after EVT is still at a high level, and the main causes of death include malignant cerebral edema, symptomatic intracranial hemorrhage and pneumonia. In recent years, with the continuous progress of EVT technology, more research focuses on the safety and effectiveness after EVT, but the systematic exploration of the cause of death after EVT is relatively insufficient. The purpose of this paper is to explain the research progress of the causes of death after EVT in acute ischemic stroke, focusing on the temporal and spatial distribution differences of the main causes of death (anterior and posterior circulation sites), and summarize other rare causes of death, so as to provide reference for optimizing clinical diagnosis and treatment strategies and reducing postoperative mortality.

Large vessel occlusive stroke (LVOS) has extremely high mortality and disability rates. Endovascular therapy is currently the most effective reperfusion therapy, but its efficacy is highly time-dependent, and delayed therapy is closely related to adverse neurological outcomes; Moreover, due to the limitations of the distribution and transport system of comprehensive stroke centers (CSCs), some LVOS patients still need to be transported step by step through primary stroke centers, resulting in delayed reperfusion treatment. Therefore, rapid and accurate prehospital identification of LVOS and direct transfer of patients to CSC is the key to shortening treatment time and improving clinical outcomes. This paper summarizes the application status and research progress of LVOS pre-hospital recognition methods.

Epileptic seizures seriously threaten the life safety of patients, and it is urgent to develop accurate seizure prediction technology to achieve early intervention. In recent years, deep learning has shown great potential in the field of seizure prediction based on EEG signals with its powerful automatic feature extraction and nonlinear modeling capabilities. This paper systematically reviews the research progress in this field in recent years, including the common methods of deep learning in seizure prediction and their technical characteristics. At the same time, the main challenges faced by the current field, such as data scarcity, insufficient model generalization capability, poor interpretability and clinical deployment, are deeply discussed. Finally, for future research directions, key paths such as transfer learning, multimodal data fusion and clinical real-time prediction system development are proposed. Through interdisciplinary cooperation, deep learning technology is expected to break through the clinical translational bottleneck of seizure prediction.