Development and validation of machine learning models to predict postoperative infarction in moyamoya disease.

OBJECTIVE: Cerebral infarction is a common complication in patients undergoing revascularization surgery for moyamoya disease (MMD). Although previous statistical evaluations have identified several risk factors for postoperative brain ischemia, the ability to predict its occurrence based on these limited predictors remains inadequately explored. This study aimed to assess the feasibility of machine learning algorithms for predicting cerebral infarction after revascularization surgery in patients with MMD. METHODS: This retrospective study was conducted across two centers and harnessed data from 512 patients with MMD who had undergone revascularization surgery. The patient cohort was partitioned into internal and external datasets. Using perioperative clinical data from the internal cohort, three distinct machine learning algorithms-namely the support vector machine, random forest, and light gradient-boosting machine models-were trained and cross-validated to predict the occurrence of postoperative cerebral infarction. Predictive performance validity was subsequently assessed using an external dataset. Shapley additive explanations (SHAP) analysis was conducted to augment the prediction model's transparency and to quantify the impact of each input variable on shaping both the aggregate and individual patient predictions. RESULTS: In the cohort of 512 patients, 33 (6.4%) experienced postrevascularization cerebral infarction. The cross-validation outcomes revealed that, among the three models, the support vector machine model achieved the largest area under the receiver operating characteristic curve (ROC-AUC) at mean ± SD 0.785 ± 0.052. Notably, during external validation, the light gradient-boosting machine model exhibited the highest accuracy at 0.903 and the largest ROC-AUC at 0.710. The top-performing prediction model utilized five input variables: postoperative serum gamma-glutamyl transpeptidase value, positive posterior cerebral artery (PCA) involvement on preoperative MRA, infarction as the rationale for surgery, presence of an infarction scar on preoperative MRI, and preoperative modified Rankin Scale score. Furthermore, the SHAP analysis identified presence of PCA involvement, infarction as the rationale for surgery, and presence of an infarction scar on preoperative MRI as positive influences on postoperative cerebral infarction. CONCLUSIONS: This study indicates the usefulness of employing machine learning techniques with routine perioperative data to predict the occurrence of cerebral infarction after revascularization procedures in patients with MMD.

Neonatal Ventricular Reservoir Implantation for Hydrocephalus Management in Chiari III Malformations: A Case Report.

Chiari III malformation, a rare and severe subtype of Chiari malformations, is frequently associated with hydrocephalus. The conventional treatment approaches for hydrocephalus in Chiari III malformations have mainly focused on ventriculoperitoneal (VP) shunting, but optimal methods and timing remain uncertain. We report a case of a newborn girl with Chiari III malformation who underwent surgical closure of an occipitocervical encephalocele and ventricular reservoir implantation on her third day of life. This procedure successfully managed her hydrocephalus without significant short-term complications. Three months post-surgery, she developed secondary external hydrocephalus, which was managed through subdural-peritoneal shunting, allowing her to thrive until at least five years of age. This case demonstrates the potential of ventricular reservoir implantation in treating hydrocephalus associated with Chiari III malformation and brings to light secondary external hydrocephalus, subsequently managed by VP shunting.

Deep learning based identification of pituitary adenoma on surgical endoscopic images: a pilot study.

Accurate tumor identification during surgical excision is necessary for neurosurgeons to determine the extent of resection without damaging the surrounding tissues. No conventional technologies have achieved reliable performance for pituitary adenomas. This study proposes a deep learning approach using intraoperative endoscopic images to discriminate pituitary adenomas from non-tumorous tissue inside the sella turcica. Static images were extracted from 50 intraoperative videos of patients with pituitary adenomas. All patients underwent endoscopic transsphenoidal surgery with a 4 K ultrahigh-definition endoscope. The tumor and non-tumorous tissue within the sella turcica were delineated on static images. Using intraoperative images, we developed and validated deep learning models to identify tumorous tissue. Model performance was evaluated using a fivefold per-patient methodology. As a proof-of-concept, the model's predictions were pathologically cross-referenced with a medical professional's diagnosis using the intraoperative images of a prospectively enrolled patient. In total, 605 static images were obtained. Among the cropped 117,223 patches, 58,088 were labeled as tumors, while the remaining 59,135 were labeled as non-tumorous tissues. The evaluation of the image dataset revealed that the wide-ResNet model had the highest accuracy of 0.768, with an F1 score of 0.766. A preliminary evaluation on one patient indicated alignment between the ground truth set by neurosurgeons, the model's predictions, and histopathological findings. Our deep learning algorithm has a positive tumor discrimination performance in intraoperative 4-K endoscopic images in patients with pituitary adenomas.

Development of machine learning models for predicting unfavorable functional outcomes from preoperative data in patients with chronic subdural hematomas.

Chronic subdural hematoma (CSDH) often causes neurological deterioration and is treated with hematoma evacuation. This study aimed to assess the feasibility of various machine learning models to preoperatively predict the functional outcome of patients with CSDH. Data were retrospectively collected from patients who underwent CSDH surgery at two institutions: one for internal validation and the other for external validation. The poor functional outcome was defined as a modified Rankin scale score of 3-6 upon hospital discharge. The unfavorable outcome was predicted using four machine learning algorithms on an internal held-out cohort (n = 188): logistic regression, support vector machine (SVM), random forest, and light gradient boosting machine. The prediction performance of these models was also validated in an external cohort (n = 99). The area under the curve of the receiver operating characteristic curve (ROC-AUC) of each machine learning-based model was found to be high in both validations (internal: 0.906-0.925, external: 0.833-0.860). In external validation, the SVM model demonstrated the highest ROC-AUC of 0.860 and accuracy of 0.919. This study revealed the potential of machine learning algorithms in predicting unfavorable outcomes at discharge among patients with CSDH undergoing burr hole surgery.

Machine learning models predict delayed hyponatremia post-transsphenoidal surgery using clinically available features.

PURPOSE: Delayed hyponatremia (DHN), a unique complication, is the leading cause of unexpected readmission after pituitary surgery. Therefore, this study aimed to develop tools for predicting postoperative DHN in patients undergoing endoscopic transsphenoidal surgery (eTSS) for pituitary neuroendocrine tumors (PitNETs). METHODS: This was a single-center, retrospective study involving 193 patients with PitNETs who underwent eTSS. The objective variable was DHN, defined as serum sodium levels < 135 mmol/L at ≥ 1 time between post operative days 3 and 9. We trained four machine learning models to predict this objective variable using the clinical variables available preoperatively and on the first postoperative day. The clinical variables included patient characteristics, pituitary-related hormone levels, blood test results, radiological findings, and postoperative complications. RESULTS: The random forest (RF) model demonstrated the highest (0.759 ± 0.039) area under the curve of the receiver operating characteristic curve (ROC-AUC), followed by the support vector machine (0.747 ± 0.034), the light gradient boosting machine (LGBM: 0.738 ± 0.026), and the logistic regression (0.710 ± 0.028). The highest accuracy (0.746 ± 0.029) was observed in the LGBM model. The best-performing RF model was based on 24 features, nine of which were clinically available preoperatively. CONCLUSIONS: The proposed machine learning models with pre- and post-resection features predicted DHN after the resection of PitNETs.

Prevention of Stroke with Closed-Cell Stent-in-Stent for Carotid Artery Stenosis Due to Stent Fracture.

Objective: Stent fracture is a risk factor for stroke. It has not been fully elucidated whether stent-in-stent procedures can effectively treat stent fractures. Case Presentation: An 80-year-old man underwent carotid artery stenting (CAS) with an open-cell stent to treat asymptomatic right internal carotid artery (ICA) stenosis. Type III stent fracture occurred during CAS. Six months later, in-stent stenosis progressed on DSA. Repeat CAS with a closed-cell stent was performed. CT showed expansion of the narrowed lumen. The patient remained stroke-free and carotid artery restenosis did not occur for 3 years postoperatively. Conclusion: Repeat CAS with a closed-cell stent is a viable treatment option for stent fracture.

Preventive Effect of Bypass Surgery on Rebleeding in Patients with Hemorrhagic Twiglike Middle Cerebral Artery.

BACKGROUND: Aplastic or twiglike middle cerebral artery (Ap/T-MCA) is a rare cerebrovascular anomaly correlated to stroke. The most common characteristic of the disease is intracerebral hemorrhage. However, treatment measures that can prevent rebleeding have not yet been established. OBJECTIVE: We sought to validate whether extracranial-intracranial (EC-IC) bypass is effective in preventing recurrent intracerebral hemorrhage in patients with Ap/T-MCA. METHODS: Ap/T-MCA was defined on the basis of the following criteria: 1) M1 occlusion with plexiform vessels, 2) unilateral MCA involvement without progression to the contralateral side, 3) no significant stenosis in the M2 segment, and 4) no stenoocclusive lesions in other intracranial major arteries. The clinical features, surgical procedures, radiologic changes, and treatment outcomes of patients with Ap/T-MCA who presented with ipsilateral intracerebral hemorrhage at our institution from April 2004 to December 2019 were retrospectively reviewed. RESULTS: Seven patients fulfilled the criteria for hemorrhagic Ap/T-MCA. Female preponderance was observed, and the mean age at the time of onset was 38 (range: 28-47) years. Three patients presented with pure intraventricular hemorrhage and 4 with intraparenchymal hemorrhage. Of the 4 patients, 2 had intraventricular extension. All patients underwent EC-IC bypass and did not present with stroke during the follow-up period (median: 36 [range: 5-140] months). Follow-up magnetic resonance imaging revealed regression of the twiglike vessels or periventricular choroidal anastomosis within 1 year after surgery in all patients. CONCLUSIONS: EC-IC bypass may be a treatment option that can prevent rebleeding in patients with hemorrhagic Ap/T-MCA.

A Rare Case of Eosinophilic Granulomatosis with Polyangiitis Presenting as Ischemic Stroke and Splenic Infarction.

BACKGROUND: Eosinophilic Granulomatosis with Polyangiitis (EGPA) is a group of systemic disorders characterized by inflammation of blood vessels and eosinophilia. Simultaneous brain and splenic infarcts are extremely rare in patients with EGPA. CASE DESCRIPTION: We report a case of a 61-year-old male with a history of asthma and sinusitis who presented with paresthesia and purpura in the lower extremities. Eosinophilia and positive Myeloperoxidase-anti-neutrophil cytoplasmic antibody were present and the diagnosis of EGPA was confirmed. Multiple bilateral cerebral and cerebellar infarcts and splenic infarction were detected. Although there was evidence of myocarditis, no cardiac thrombus was detected. Immunosuppressive and anticoagulation therapy were provided. The patient was fully recovered. CONCLUSIONS: EGPA can present as splenic infarction and ischemic stroke. Prompt diagnosis and treatment with anticoagulant and immunosuppressive agents may lead to good prognosis.

A Case of Mechanical Thrombectomy for Acute Occlusion of the Left Internal Carotid Artery Later than 24 Hours after Onset.

Objective: Mechanical thrombectomy is performed on ischemic stroke patients with acute major cerebral artery occlusion within 24 hours of symptom onset. We report a case of delayed mechanical thrombectomy for acute left internal carotid artery occlusion. Case Presentation: A 76-year-old woman suddenly presented with dysarthria and right hemiparesis was admitted to her previous hospital. She was treated by conservative therapy. The next day, she was transferred to our hospital 26 hours after onset with a diagnosis of ischemic stroke due to left carotid artery acute occlusion. Contrast CT revealed left carotid artery occlusion. Arterial fibrillation was detected. Mechanical thrombectomy through the right brachial artery was immediately performed. Complete recanalization was achieved without hemorrhagic complication. Her postoperative course was uneventful. Conclusion: In this case, delayed mechanical thrombectomy for acute major cerebral artery occlusion was safely performed 24 hours after symptom onset.

Non-brain-exposure Exoscopic and Endoscopic Volume Reduction Surgery for Benign Meningioma En Plaque in an Elderly Patient: A Case Report.

Meningioma en plaque (MEP) often needs invasive surgery to resect totally because of its widespread proliferation along the dura mater. We report a minimally invasive surgical technique for non-brain-exposure exoscopic and endoscopic volume reduction in an elderly patient with MEP. An 83-year-old woman presented with gait disturbance and cognitive dysfunction that had progressed over 6 months. Magnetic resonance imaging (MRI) revealed a large MEP on the right frontal lobe with peritumoral edema. On confirming the benign nature of the tumor (WHO grade 1) by biopsy, the main feeders and tumor were transarterially embolized. Volume reduction surgery was performed under the assistance of an exoscope and an endoscope while being careful not to expose and damage the cortex. Her symptoms completely resolved postoperatively. This surgical technique without exposing the brain may be a treatment choice for elderly patients with benign symptomatic convexity MEP.

Long-term Outcomes of Carotid Artery Stenting: A Single-center Experience.

The long-term prophylactic effect of carotid artery stenting (CAS) remains incompletely elucidated. We evaluated outcomes of CAS at our institution to determine the safety and efficacy of CAS in real-world settings. We retrospectively analyzed 73 patients who underwent CAS from 2006 to 2013. Periprocedural results were compared between asymptomatic and symptomatic carotid stenosis groups. The primary endpoint was a composite of ipsilateral stroke, death, and carotid artery restenosis beyond 30 days and within 5 years after the first procedure. The average age was 72.2 years with a majority of male subjects (84.9%). Twenty-seven patients (37%) were asymptomatic. Incidence of periprocedural adverse events and mRS ≤2 at 30 days after CAS were not significantly different between groups (P = 0.14 and 0.07, respectively). CAS was unsuccessful in three patients and one post-procedural minor stroke occurred. Therefore, 69 patients were included in the long-term study. The rate of occurrence of the primary endpoint was 21.7%. Ipsilateral ischemic stroke occurred in one patient, which was due to cardiogenic embolus. Nine patients died, and cancer was the most frequent cause. Five in-stent restenoses were observed. All patients with restenosis underwent additional CAS without any occurrence of stroke. This study revealed the safety and long-term efficacy of CAS in a real-world setting. Routine follow-up is also important for detecting carotid artery restenosis.

Long-Term Effects on Preventing Stroke after Endovascular Treatment or Bypass Surgery for Intracranial Arterial Stenosis.

BACKGROUND: Intracranial arterial stenosis (ICAS) is an important cause of ischemic stroke worldwide due to its higher risk of recurrence with medical therapy. Although some large randomized studies failed to show the superiority of surgical treatment compared with medical therapy, the results of medical therapy are not sufficient. There are patients who still benefit from surgical treatment. This retrospective analysis aimed to evaluate the long-term efficacy of surgical therapy with percutaneous transluminal angioplasty and/or stenting (PTA/PTAS) or extracranial-intracranial (EC/IC) bypass surgery for patients with ICAS. METHODS: Between October 2005 and December 2016, 55 ICAS patients were treated with PTA/PTAS or EC-IC bypass surgery. Their electronic medical records were retrospectively reviewed and analyzed. The primary outcome was all adverse events beyond 30 days after a revascularization procedure. RESULTS: We performed 21 cases (35%) of PTA, 4 cases (7%) of PTAS, and 34 cases (58%) of EC-IC bypass surgery and the median follow-up duration was 66 months (range 1-144 months). The occurrence rate of the primary outcome was 10.2% and only 1 patient (1.8%) experienced ipsilateral disabling ischemic stroke beyond 30 days. The long-term functional independent survival rate was 83.6%. CONCLUSIONS: We demonstrated a long-term favorable outcome of combined surgical intervention for ICAS patients with PTA/PTAS and EC-IC bypass surgery, and the result was better than previously reported outcomes of medical therapy. Additional multicenter studies are required to draw firm conclusions on the efficacy of reduction of recurrent stroke in patients with ICAS.