Detecting Cerebral Ischemia From Electroencephalography During Carotid Endarterectomy Using Machine Learning.

Monitoring cerebral neuronal activity via electroencephalography (EEG) during surgery can detect ischemia, a precursor to stroke. However, current neurophysiologist-based monitoring is prone to error. In this study, we evaluated machine learning (ML) for efficient and accurate ischemia detection. We trained supervised ML models on a dataset of 802 patients with intraoperative ischemia labels and evaluated them on an independent validation dataset of 30 patients with refined labels from five neurophysiologists. Our results show moderate-to-substantial agreement between neurophysiologists, with Cohen's kappa values between 0.59 and 0.74. Neurophysiologist performance ranged from 58-93% for sensitivity and 83-96% for specificity, while ML models demonstrated comparable ranges of 63-89% and 85-96%. Random Forest (RF), LightGBM (LGBM), and XGBoost RF achieved area under the receiver operating characteristic curve (AUROC) values of 0.92-0.93 and area under the precision-recall curve (AUPRC) values of 0.79-0.83. ML has the potential to improve intraoperative monitoring, enhancing patient safety and reducing costs.

Time-Series Aware Metrics for the Evaluation of Intraoperative Electroencephalography-Based Ischemia Detection.

Continuous intraoperative monitoring with electroencephalo2 graphy (EEG) is commonly used to detect cerebral ischemia in high-risk surgical procedures such as carotid endarterectomy. Machine learning (ML) models that detect ischemia in real time can form the basis of automated intraoperative EEG monitoring. In this study, we describe and compare two time-series aware precision and recall metrics to the classical precision and recall metrics for evaluating the performance of ML models that detect ischemia. We trained six ML models to detect ischemia in intraoperative EEG and evaluated them with the area under the precision-recall curve (AUPRC) using time-series aware and classical approaches to compute precision and recall. The Support Vector Classification (SVC) model performed the best on the time-series aware metrics, while the Light Gradient Boosting Machine (LGBM) model performed the best on the classical metrics. Visual inspection of the probability outputs of the models alongside the actual ischemic periods revealed that the time-series aware AUPRC selected a model more likely to predict ischemia onset in a timely fashion than the model selected by classical AUPRC.

Effect of hospitalization on gastrointestinal motility and pH in dogs.

OBJECTIVE To determine the effect of hospitalization on gastrointestinal motility and pH in healthy dogs. DESIGN Experimental study. ANIMALS 12 healthy adult dogs. PROCEDURES A wireless motility capsule (WMC) that measured pressure, transit time, and pH within the gastrointestinal tract was administered orally to dogs in 2 phases. In the first phase, dogs received the WMC at the hospital and then returned to their home to follow their daily routine. In the second phase, dogs were hospitalized, housed individually, had abdominal radiography performed daily, and were leash exercised 4 to 6 times/d until the WMC passed in the feces. All dogs received the same diet twice per day in both phases. Data were compared between phases with the Wilcoxon signed rank test. RESULTS Data were collected from 11 dogs; 1 dog was excluded because the WMC failed to exit the stomach. Median gastric emptying time during hospitalization (71.8 hours; range, 10.7 to 163.0 hours) was significantly longer than at home (17.6 hours; range, 9.7 to 80.8 hours). Values of all other gastric, small bowel, and large bowel parameters (motility index, motility pattern, pH, and transit time) were similar between phases. No change in gastric pH was detected over the hospitalization period. High interdog variability was evident for all measured parameters. CONCLUSIONS AND CLINICAL RELEVANCE Hospitalization of dogs may result in a prolonged gastric emptying time, which could adversely affect gastric emptying of meals, transit of orally administered drugs, or assessments of underlying motility disorders.

Intravesical glycosaminoglycans for obstructive feline idiopathic cystitis: a pilot study.

Feline idiopathic cystitis is a common condition, often resulting in repeated episodes of life-threatening urethral obstruction. Defective urinary bladder glycosaminoglycans have been implicated as a causal factor. In this report, a commercially available glycosaminoglycan product was infused into the urinary bladders of cats with urethral obstruction from idiopathic cystitis to study the effect on repeated obstruction. In this randomized, blind, placebo-controlled clinical trial, the therapeutic protocol was well tolerated with no adverse effects. Whereas no glycosaminoglycan-treated cats (n = 9) developed repeated urethral obstruction during the 7 day follow-up period, 3/7 placebo-treated cats developed repeated obstructions. Approaching statistical significance (P = 0.06), these data suggest that further investigation of this new treatment option is warranted.