Predicting respiratory decompensation in mechanically ventilated adult ICU patients.

Introduction: Mechanical ventilation is a life-saving treatment in the Intensive Care Unit (ICU), but often causes patients to be at risk of further respiratory complication. We created a statistical model utilizing electronic health record and physiologic vitals data to predict the Center for Disease Control and Prevention (CDC) defined Ventilator Associated Complications (VACs). Further, we evaluated the effect of data temporal resolution and feature generation method choice on the accuracy of such a constructed model. Methods: We constructed a random forest model to predict occurrence of VACs using health records and chart events from adult patients in the Medical Information Mart for Intensive Care III (MIMIC-III) database. We trained the machine learning models on two patient populations of 1921 and 464 based on low and high frequency data availability. Model features were generated using both basic statistical summaries and tsfresh, a python library that generates a large number of derived time-series features. Classification to determine whether a patient will experience VAC one hour after 35 h of ventilation was performed using a random forest classifier. Two different sample spaces conditioned on five varying feature extraction techniques were evaluated to identify the most optimal selection of features resulting in the best VAC discrimination. Each dataset was assessed using K-folds cross-validation (k = 10), giving average area under the receiver operating characteristic curves (AUROCs) and accuracies. Results: After feature selection, hyperparameter tuning, and feature extraction, the best performing model used automatically generated features on high frequency data and achieved an average AUROC of 0.83 ± 0.11 and an average accuracy of 0.69 ± 0.10. Discussion: Results show the potential viability of predicting VACs using machine learning, and indicate that higher-resolution data and the larger feature set generated by tsfresh yield better AUROCs compared to lower-resolution data and manual statistical features.

Should asymptomatic cervical stenosis be treated in the setting of progressive thoracic myelopathy? A systematic review of the literature.

PURPOSE: Unlike tandem stenosis of the cervical and lumbar spine, tandem cervical and thoracic stenosis (TCTS) of the spine is less common, and the approach and order of intervention are controversial. We aim to review the literature to evaluate the incidence and interventions for patients with cervical and thoracic stenosis. We provide illustrative cases to demonstrate that thoracic myelopathy in the setting of asymptomatic cervical stenosis can be treated safely. METHODS: A systematic review of the literature through electronic databases of PubMed, EMBASE, Web of Science, and Cochrane Library was performed to present the current literature that evaluates TCTS as it relates to incidence and surgical interventions. We also present two cases of patients undergoing operative intervention for thoracic myelopathy in the setting of concurrent cervical stenosis. RESULTS: A total of 26 English original studies and case reports were identified. Nine studies evaluated the incidence of TCTS. 20 studies with a total of 168 patients with TCTS presented information on surgical intervention options. There is an overall aggregate incidence of 11.6% (530/4751) based on incidence studies. 165 patients underwent thoracic intervention. Of these patients, 63 patients underwent cervical intervention first, 29 underwent thoracic intervention first, and 73 underwent simultaneous, single-stage intervention. CONCLUSIONS: In patients presenting with myelopathy, both cervical and thoracic spine should be evaluated for TCTS. Order of operative intervention is tailored to clinical and radiographic information. In cases of thoracic myelopathy with asymptomatic cervical stenosis, thoracic intervention can be pursued with precautions to prevent further cervical cord injury.

Early Detection of Hypothermic Neuroprotection Using T2-Weighted Magnetic Resonance Imaging in a Mouse Model of Hypoxic Ischemic Encephalopathy.

Perinatal hypoxic-ischemic encephalopathy (HIE) can lead to neurodevelopmental disorders, including cerebral palsy. Standard care for neonatal HIE includes therapeutic hypothermia, which provides partial neuroprotection; magnetic resonance imaging (MRI) is often used to assess injury and predict outcome after HIE. Immature rodent models of HIE are used to evaluate mechanisms of injury and to examine the efficacy and mechanisms of neuroprotective interventions such as hypothermia. In this study, we first confirmed that, in the CD1 mouse model of perinatal HIE used for our research, MRI obtained 3 h after hypoxic ischemia (HI) could reliably assess initial brain injury and predict histopathological outcome. Mice were subjected to HI (unilateral carotid ligation followed by exposure to hypoxia) on postnatal day 7 and were imaged with T2-weighted MRI and diffusion-weighted MRI (DWI), 3 h after HI. Clearly defined regions of increased signal were comparable in T2 MRI and DWI, and we found a strong correlation between T2 MRI injury scores 3 h after HI and histopathological brain injury 7 days after HI, validating this method for evaluating initial injury in this model of HIE. The more efficient, higher resolution T2 MRI was used to score initial brain injury in subsequent studies. In mice treated with hypothermia, we found a significant reduction in T2 MRI injury scores 3 h after HI, compared to normothermic littermates. Early hypothermic neuroprotection was maintained 7 days after HI, in both T2 MRI injury scores and histopathology. In the normothermic group, T2 MRI injury scores 3 h after HI were comparable to those obtained 7 days after HI. However, in the hypothermic group, brain injury was significantly less 7 days after HI than at 3 h. Thus, early neuroprotective effects of hypothermia were enhanced by 7 days, which may reflect the additional 3 h of hypothermia after imaging or effects on later mechanisms of injury, such as delayed cell death and inflammation. Our results demonstrate that hypothermia has early neuroprotective effects in this model. These findings suggest that hypothermia has an impact on early mechanisms of excitotoxic injury and support initiation of hypothermic intervention as soon as possible after diagnosis of HIE.