Predicting outcome after aneurysmal subarachnoid hemorrhage by exploitation of signal complexity: a prospective two-center cohort study.

BACKGROUND: Signal complexity (i.e. entropy) describes the level of order within a system. Low physiological signal complexity predicts unfavorable outcome in a variety of diseases and is assumed to reflect increased rigidity of the cardio/cerebrovascular system leading to (or reflecting) autoregulation failure. Aneurysmal subarachnoid hemorrhage (aSAH) is followed by a cascade of complex systemic and cerebral sequelae. In aSAH, the value of entropy has not been established yet. METHODS: aSAH patients from 2 prospective cohorts (Zurich-derivation cohort, Aachen-validation cohort) were included. Multiscale Entropy (MSE) was estimated for arterial blood pressure, intracranial pressure, heart rate, and their derivatives, and compared to dichotomized (1-4 vs. 5-8) or ordinal outcome (GOSE-extended Glasgow Outcome Scale) at 12 months using uni- and multivariable (adjusted for age, World Federation of Neurological Surgeons grade, modified Fisher (mFisher) grade, delayed cerebral infarction), and ordinal methods (proportional odds logistic regression/sliding dichotomy). The multivariable logistic regression models were validated internally using bootstrapping and externally by assessing the calibration and discrimination. RESULTS: A total of 330 (derivation: 241, validation: 89) aSAH patients were analyzed. Decreasing MSE was associated with a higher likelihood of unfavorable outcome independent of covariates and analysis method. The multivariable adjusted logistic regression models were well calibrated and only showed a slight decrease in discrimination when assessed in the validation cohort. The ordinal analysis revealed its effect to be linear. MSE remained valid when adjusting the outcome definition against the initial severity. CONCLUSIONS: MSE metrics and thereby complexity of physiological signals are independent, internally and externally valid predictors of 12-month outcome. Incorporating high-frequency physiological data as part of clinical outcome prediction may enable precise, individualized outcome prediction. The results of this study warrant further investigation into the cause of the resulting complexity as well as its association to important and potentially preventable complications including vasospasm and delayed cerebral ischemia.

Thrombocyte transfusion and rebleeding rate in patients using antiplatelet agents before aneurysmal subarachnoid hemorrhage.

OBJECTIVE: The reason for a rebleed after an initial hemorrhage in patients with aneurysmal subarachnoid hemorrhage (aSAH) is considered multifactorial. Antiplatelet use is one of the factors that has been related to early rebleed and worse outcome after aSAH. Thrombocyte transfusion overcomes the inhibitory effects of antiplatelet agents by increasing the number of functional thrombocytes, but its impact on the rebleed rate and clinical outcome remains unknown. The aim of this study was to assess the effect of thrombocyte transfusion on rebleeding and clinical outcome in patients with aSAH and prehemorrhage antiplatelet use, considering confounding factors. METHODS: Data were prospectively collected at a single tertiary reference center for aSAH in Zurich, Switzerland. Patients with aSAH and prehemorrhage antiplatelet use were divided into "thrombocyte transfusion" and "nontransfusion" groups based on whether they did or did not receive any thrombocyte transfusion in the acute stage of aSAH after hospital admission and before the exclusion of the bleeding source. Using multivariate logistic regression analysis, the impact of thrombocyte transfusion on the rebleed rate and on clinical outcome (defined as Glasgow Outcome Scale score 1-3) was calculated. RESULTS: One hundred fifty-seven patients were included, 87 (55.4%) of whom received thrombocyte transfusion. Eighteen (11.5%) of 157 patients had a rebleed during the hospital stay. The rebleed risk was 6.9% in the thrombocyte transfusion group and 17.1% in the nontransfusion group. After adjusting for confounders, thrombocyte transfusion showed evidence for a reduction in the rebleed rate (adjusted OR [aOR] 0.29, 95% CI 0.10-0.87). Fifty-seven patients (36.3%) achieved a poor outcome at 6 months' follow-up. Among those 57 patients, 31 (54.4%) underwent at least one thrombocyte transfusion. Thrombocyte transfusion was not associated with poor clinical outcome at 6 months' follow-up (aOR 0.91, 95% CI 0.39-2.15). CONCLUSIONS: Thrombocyte transfusion in patients with aSAH and prehemorrhage antiplatelet use is independently associated with a reduction in rebleeds but shows no impact on clinical outcome at 6 months' follow-up. Larger and randomized studies are needed to investigate the impact of thrombocyte transfusion on rebleed and outcome.

Spontaneous empyema and brain abscess in an intensive care population: clinical presentation, microbiology, and factors associated with outcome.

BACKGROUND: Data on critically ill patients with spontaneous empyema or brain abscess are limited. The aim was to evaluate clinical presentations, factors, and microbiological findings associated with the outcome in patients treated in a Neurocritical Care Unit. METHODS: In this retrospective study, we analyzed 45 out of 101 screened patients with spontaneous epidural or subdural empyema and/or brain abscess treated at a tertiary care center between January 2012 and December 2019. Patients with postoperative infections or spinal abscess were excluded. Medical records were reviewed for baseline characteristics, origin of infection, laboratory and microbiology findings, and treatment characteristics. The outcome was determined using the Glasgow outcome scale extended (GOSE). RESULTS: Favorable outcome (GOSE 5-8) was achieved in 38 of 45 patients (84%). Four patients died (9%), three remained severely disabled (7%). Unfavorable outcome was associated with a decreased level of consciousness at admission (Glasgow coma scale < 9) (43% versus 3%; p = 0.009), need of vasopressors (71% versus 11%; p = 0.002), sepsis (43% versus 8%; p = 0.013), higher age (65.1 ± 15.7 versus 46.9 ± 17.5 years; p = 0.014), shorter time between symptoms onset and ICU admission (5 ± 2.4 days versus 11.6 ± 16.8 days; p = 0.013), and higher median C-reactive protein (CRP) serum levels (206 mg/l, range 15-259 mg/l versus 17.5 mg/l, range 3.3-72.7 mg/l; p = 0.036). With antibiotics adapted according to culture sensitivities in the first 2 weeks, neuroimaging revealed a progression of empyema or abscess in 45% of the cases. CONCLUSION: Favorable outcome can be achieved in a considerable proportion of an intensive care population with spontaneous empyema or brain abscess. Sepsis and more frequent need for vasopressors, associated with unfavorable outcome, indicate a fulminant course of a not only cerebral but systemic infection. Change of antibiotic therapy according to microbiological findings in the first 2 weeks should be exercised with great caution.

Cerebrospinal fluid findings in COVID-19: a multicenter study of 150 lumbar punctures in 127 patients.

BACKGROUND: Comprehensive data on the cerebrospinal fluid (CSF) profile in patients with COVID-19 and neurological involvement from large-scale multicenter studies are missing so far. OBJECTIVE: To analyze systematically the CSF profile in COVID-19. METHODS: Retrospective analysis of 150 lumbar punctures in 127 patients with PCR-proven COVID-19 and neurological symptoms seen at 17 European university centers RESULTS: The most frequent pathological finding was blood-CSF barrier (BCB) dysfunction (median QAlb 11.4 [6.72-50.8]), which was present in 58/116 (50%) samples from patients without pre-/coexisting CNS diseases (group I). QAlb remained elevated > 14d (47.6%) and even > 30d (55.6%) after neurological onset. CSF total protein was elevated in 54/118 (45.8%) samples (median 65.35 mg/dl [45.3-240.4]) and strongly correlated with QAlb. The CSF white cell count (WCC) was increased in 14/128 (11%) samples (mostly lympho-monocytic; median 10 cells/µl, > 100 in only 4). An albuminocytological dissociation (ACD) was found in 43/115 (37.4%) samples. CSF L-lactate was increased in 26/109 (24%; median 3.04 mmol/l [2.2-4]). CSF-IgG was elevated in 50/100 (50%), but was of peripheral origin, since QIgG was normal in almost all cases, as were QIgA and QIgM. In 58/103 samples (56%) pattern 4 oligoclonal bands (OCB) compatible with systemic inflammation were present, while CSF-restricted OCB were found in only 2/103 (1.9%). SARS-CoV-2-CSF-PCR was negative in 76/76 samples. Routine CSF findings were normal in 35%. Cytokine levels were frequently elevated in the CSF (often associated with BCB dysfunction) and serum, partly remaining positive at high levels for weeks/months (939 tests). Of note, a positive SARS-CoV-2-IgG-antibody index (AI) was found in 2/19 (10.5%) patients which was associated with unusually high WCC in both of them and a strongly increased interleukin-6 (IL-6) index in one (not tested in the other). Anti-neuronal/anti-glial autoantibodies were mostly absent in the CSF and serum (1509 tests). In samples from patients with pre-/coexisting CNS disorders (group II [N = 19]; including multiple sclerosis, JC-virus-associated immune reconstitution inflammatory syndrome, HSV/VZV encephalitis/meningitis, CNS lymphoma, anti-Yo syndrome, subarachnoid hemorrhage), CSF findings were mostly representative of the respective disease. CONCLUSIONS: The CSF profile in COVID-19 with neurological symptoms is mainly characterized by BCB disruption in the absence of intrathecal inflammation, compatible with cerebrospinal endotheliopathy. Persistent BCB dysfunction and elevated cytokine levels may contribute to both acute symptoms and 'long COVID'. Direct infection of the CNS with SARS-CoV-2, if occurring at all, seems to be rare. Broad differential diagnostic considerations are recommended to avoid misinterpretation of treatable coexisting neurological disorders as complications of COVID-19.

Blood Pressure Variability Indices for Outcome Prediction After Thrombectomy in Stroke by Using High-Resolution Data.

BACKGROUND: Blood pressure variability (BPV) is associated with outcome after endovascular thrombectomy in acute large vessel occlusion stroke. We aimed to provide the optimal sampling frequency and BPV index for outcome prediction by using high-resolution blood pressure (BP) data. METHODS: Patient characteristics, 3-month outcome, and BP values measured intraarterially at 1 Hz for up to 24 h were extracted from 34 patients treated at a tertiary care center neurocritical care unit. Outcome was dichotomized (modified Rankin Scale 0-2, favorable, and 3-6, unfavorable) and associated with systolic BPV (as calculated by using standard deviation, coefficient of variation, averaged real variability, successive variation, number of trend changes, and a spectral approach using the power of specific BP frequencies). BP values were downsampled by either averaging or omitting all BP values within each prespecified time bin to compare the different sampling rates. RESULTS: Out of 34 patients (age 72 ± 12.7 years, 67.6% men), 10 (29.4%) achieved a favorable functional outcome and 24 (70.6%) had an unfavorable functional outcome at 3 months. No group differences were found in mean absolute systolic BP (SBP) (130 ± 18 mm Hg, p = 0.82) and diastolic BP (DBP) (59 ± 10 mm Hg, p = 1.00) during the monitoring time. BPV only reached predictive significance when using successive variation extracted from downsampled (averaged over 5 min) SBP data (median 4.8 mm Hg [range 3.8-7.1]) in patients with favorable versus 7.1 mmHg [range 5.5-9.7] in those with unfavorable outcome, area under the curve = 0.74 [confidence interval (CI) 0.57-0.85; p = 0.031], or the power of midrange frequencies between 1/20 and 1/5 min [area under the curve = 0.75 (CI 0.59-0.86), p = 0.020]. CONCLUSIONS: Using high-resolution BP data of 1 Hz, downsampling by averaging all BP values within 5-min intervals is essential to find relevant differences in systolic BPV, as noise can be avoided (confirmed by the significance of the power of midrange frequencies). These results demonstrate how high-resolution BP data can be processed for effective outcome prediction.

The effect of nimodipine on pulmonary function in artificially ventilated patients with aneurysmal subarachnoid hemorrhage.

BACKGROUND: Nimodipine is routinely administered in patients with aneurysmal subarachnoid hemorrhage (aSAH). However, the effect of nimodipine on oxygen exchange in the lungs is insufficiently explored. METHODS: The study explored nimodipine medication in artificially ventilated patients with aSAH. The data collection period was divided into nimodipine-dependent (ND) and nimodipine-independent (NID) periods. Values for arterial partial pressure of oxygen (PaO2) and fraction of inspired oxygen (FiO2) were collected and compared between the periods. Patients were divided in those with lung injury (LI), defined as median Horowitz index (PaO2/FiO2) ≤40 kPa (≤300 mmHg), and without and in those with lower respiratory tract infection (LRTI) and without. RESULTS: A total of 53 out of 150 patients were artificially ventilated, and in 29 patients, the Horowitz index could be compared between ND and NID periods. A linear mixed model showed that during ND period the Horowitz index was 2.3 kPa (95% CI, 1.0-3.5 kPa, P<0.001) lower when compared to NID period. The model suggested that in the presence of LI, ND period is associated with a decrease of the index by 2.8 kPa (95% CI, 1.2-4.3 kPa, P<0.001). The decrease was more pronounced with LRTI than without: 3.4 kPa (95% CI, 0.8-6.1 kPa) vs. 2.1 kPa (95% CI, 0.7-3.4 kPa), P=0.011 and P=0.002, respectively. CONCLUSIONS: In patients with LI or LRTI in the context of aSAH, pulmonary function may worsen with nimodipine treatment. The drop of 2 to 3 kPa of the Horowitz index in patients with no lung pathology may not outweigh the benefits of nimodipine. However, in individuals with concomitant lung injury, the effect may be clinically relevant.

Large and Small Cerebral Vessel Involvement in Severe COVID-19: Detailed Clinical Workup of a Case Series.

BACKGROUND AND PURPOSE: Case series indicating cerebrovascular disorders in coronavirus disease 2019 (COVID-19) have been published. Comprehensive workups, including clinical characteristics, laboratory, electroencephalography, neuroimaging, and cerebrospinal fluid findings, are needed to understand the mechanisms. METHODS: We evaluated 32 consecutive critically ill patients with COVID-19 treated at a tertiary care center from March 9 to April 3, 2020, for concomitant severe central nervous system involvement. Patients identified underwent computed tomography, magnetic resonance imaging, electroencephalography, cerebrospinal fluid analysis, and autopsy in case of death. RESULTS: Of 32 critically ill patients with COVID-19, 8 (25%) had severe central nervous system involvement. Two presented with lacunar ischemic stroke in the early phase and 6 with prolonged impaired consciousness after termination of analgosedation. In all but one with delayed wake-up, neuroimaging or autopsy showed multiple cerebral microbleeds, in 3 with additional subarachnoid hemorrhage and in 2 with additional small ischemic lesions. In 3 patients, intracranial vessel wall sequence magnetic resonance imaging was performed for the first time to our knowledge. All showed contrast enhancement of vessel walls in large cerebral arteries, suggesting vascular wall pathologies with an inflammatory component. Reverse transcription-polymerase chain reactions for SARS-CoV-2 in cerebrospinal fluid were all negative. No intrathecal SARS-CoV-2-specific IgG synthesis was detectable. CONCLUSIONS: Different mechanisms of cerebrovascular disorders might be involved in COVID-19. Acute ischemic stroke might occur early. In a later phase, microinfarctions and vessel wall contrast enhancement occur, indicating small and large cerebral vessels involvement. Central nervous system disorders associated with COVID-19 may lead to long-term disabilities. Mechanisms should be urgently investigated to develop neuroprotective strategies.

Missed Anterior Inferior Cerebellar Artery Aneurysm Mimicking Vestibular Neuritis-Clues to Prevent Misdiagnosis.

We discuss a case with combined vestibulocochlear and facial neuropathy mimicking a less urgent peripheral vestibular pattern of acute vestibular syndrome (AVS). With initial magnetic resonance imaging read as normal, the patient was treated for vestibular neuropathy until headaches worsened and a diagnosis of subarachnoid hemorrhage was made. On conventional angiography, a ruptured distal right-sided aneurysm of the anterior inferior cerebellar artery was diagnosed and coiled. Whereas acute vestibular loss usually points to a benign peripheral cause of AVS, combined neuropathy of the vestibulocochlear and the facial nerve requires immediate neuroimaging focusing on the cerebellopontine angle. Imaging should be assessed jointly by neuroradiologists and the clinicians in charge to take the clinical context into account.

Low frequency oscillations reflect neurovascular coupling and disappear after cerebral death.

Spectrum power analysis in the low frequency oscillations (LFO) region of functional near infrared spectroscopy (fNIRS) is a promising method to deliver information about brain activation and therefore might be used for prognostication in patients with disorders of consciousness in the neurocritical care unit alongside with established methods. In this study, we measure the cortical hemodynamic response measured by fNIRS in the LFO region following auditory and somatosensory stimulation in healthy subjects. The significant hemodynamic reaction in the contralateral hemisphere correlation with the physiologic electric response suggests neurovascular coupling. In addition, we investigate power spectrum changes in steady state measurements of cerebral death patients and healthy subjects in the LFO region, the frequency of the heartbeat and respiration. The spectral power within the LFO region was lower in the patients with cerebral death compared to the healthy subjects, whereas there were no differences in spectral power for physiological activities such as heartbeat and respiration rate. This finding indicates the cerebral origin of our low frequency measurements. Therefore, LFO measurements are a potential method to detect brain activation in patients with disorders of consciousness and cerebral death. However, further studies in patients are needed to investigate its potential clinical use.

A Longitudinal Model for a Dynamic Risk Score to Predict Delayed Cerebral Ischemia after Subarachnoid Hemorrhage.

BACKGROUND: Accurate longitudinal risk prediction for DCI (delayed cerebral ischemia) occurrence after subarachnoid hemorrhage (SAH) is essential for clinicians to administer appropriate and timely diagnostics, thereby improving treatment planning and outcome. This study aimed to develop an improved longitudinal DCI prediction model and evaluate its performance in predicting DCI between day 4 and 14 after aneurysm rupture. METHODS: Two DCI classification models were trained: (1) a static model based on routinely collected demographics and SAH grading scores and (2) a dynamic model based on results from laboratory and blood gas analysis anchored at the time of DCI. A combined model was derived from these two using a voting approach. Multiple classifiers, including Logistic Regression, Support Vector Machines, Random Forests, Histogram-based Gradient Boosting, and Extremely Randomized Trees, were evaluated through cross-validation using anchored data. A leave-one-out simulation was then performed on the best-performing models to evaluate their longitudinal performance using time-dependent Receiver Operating Characteristic (ROC) analysis. RESULTS: The training dataset included 218 patients, with 89 of them developing DCI (41%). In the anchored ROC analysis, the combined model achieved a ROC AUC of 0.73 ± 0.05 in predicting DCI onset, the static and the dynamic model achieved a ROC AUC of 0.69 ± 0.08 and 0.66 ± 0.08, respectively. In the leave-one-out simulation experiments, the dynamic and voting model showed a highly dynamic risk score (intra-patient score range was 0.25 [0.24, 0.49] and 0.17 [0.12, 0.25] for the dynamic and the voting model, respectively, for DCI occurrence over the course of disease. In the time-dependent ROC analysis, the dynamic model performed best until day 5.4, and afterwards the voting model showed the best performance. CONCLUSIONS: A machine learning model for longitudinal DCI risk assessment was developed comprising a static and a dynamic sub-model. The longitudinal performance evaluation highlighted substantial time dependence in model performance, underscoring the need for a longitudinal assessment of prediction models in intensive care settings. Moreover, clinicians need to be aware of these performance variations when performing a risk assessment and weight the different model outputs correspondingly.

Evaluation of machine learning algorithms for noninvasive intracranial pressure estimation using near infrared spectroscopy as a covariate.

BACKGROUND: Intracranial pressure (ICP) is a vital parameter that is continuously monitored in patients with severe brain injury and imminent intracranial hypertension. OBJECTIVE: To estimate intracranial pressure without intracranial probes based on transcutaneous near infrared spectroscopy (NIRS). METHODS: We developed machine learning based approaches for noninvasive intracranial pressure (ICP) estimation using signals from transcutaneous near infrared spectroscopy (NIRS) as well as other cardiovascular and artificial ventilation parameters. RESULTS: In a patient cohort of 25 patients, with 22 used for model development and 3 for model testing, the best performing models were Fourier transform based Transformer ICP waveform estimation which produced a mean absolute error of 4.68 mm Hg (SD = 5.4) in estimation. CONCLUSION: We did not find a significant improvement in ICP estimation accuracy by including signals measured by transcutaneous NIRS. We expect that with higher quality and greater volume of data, noninvasive estimation of ICP will improve.

Survey-based qualitative exploration of user perspectives on the philips visual patient avatar in clinical situation management.

Philips Visual Patient Avatar is an innovative approach to patient monitoring. Computer-based simulation studies have shown that it can improve diagnostic accuracy and confidence while reducing perceived workload. Following its integration into clinical practice, we conducted a single-centre qualitative study at the University Hospital Zurich to explore the views of anaesthesia, post-anaesthesia and intensive care providers on their experience with the technology. We used an online survey to assess its contributions in different clinical situations. We analysed the data thematically to identify key themes. Of the 510 healthcare providers contacted, 131 (25.7%) completed the survey and 154 comments were collected. Key themes included the detection of specific vital sign changes, focusing on temperature and oxygen saturation (41.9%, 34/81 comments in the operating room; 38.6%, 17/44 comments in the intensive care unit; 10.3%, 3/29 comments in the post-anaesthesia care unit). Additionally, the technology was perceived to support daily routines and situational awareness (28.4%, 23/81 comments in the OR; 9.1%, 4/44 comments in the ICU; 10.3%, 3/29 comments in the PACU). The study provides early, but strong evidence that the Philips Visual Patient Avatar assists healthcare providers in specific clinical situations in the perioperative and critical care settings.

Intraparenchymal near-infrared spectroscopy for detection of delayed cerebral ischemia in poor-grade aneurysmal subarachnoid hemorrhage.

OBJECTIVE: Detection of delayed cerebral ischemia (DCI) is challenging in comatose patients with poor-grade aneurysmal subarachnoid hemorrhage (aSAH). Brain tissue oxygen pressure (PbtO2) monitoring may allow early detection of its occurrence. Recently, a probe for combined measurement of intracranial pressure (ICP) and intraparenchymal near-infrared spectroscopy (NIRS) has become available. In this pilot study, the parameters PbtO2, Hboxy, Hbdeoxy, Hbtotal and rSO2 were measured in parallel and evaluated for their potential to detect perfusion deficits or cerebral infarction. METHODS: In patients undergoing multimodal neuromonitoring due to poor neurological condition after aSAH, Clark oxygen probes, microdialysis and NIRS-ICP probes were applied. DCI was suspected when the measured parameters in neuromonitoring deteriorated. Thus, perfusion CT scan was performed as follow up, and DCI was confirmed as perfusion deficit. Median values for PbtO2, Hboxy, Hbdeoxy, Hbtotal and rSO2 in patients with perfusion deficit (Tmax > 6 s in at least 1 vascular territory) and/or already demarked infarcts were compared in 24- and 48-hour time frames before imaging. RESULTS: Data from 19 patients (14 University Hospital Zurich, 5 Charité Universitätsmedizin Berlin) were prospectively collected and analyzed. In patients with perfusion deficits, the median values for Hbtotal and Hboxy in both time frames were significantly lower. With perfusion deficits, the median values for Hboxy and Hbtotal in the 24 h time frame were 46,3 [39.6, 51.8] µmol/l (no perfusion deficits 53 [45.9, 55.4] µmol/l, p = 0.019) and 69,3 [61.9, 73.6] µmol/l (no perfusion deficits 74,6 [70.1, 79.6] µmol/l, p = 0.010), in the 48 h time frame 45,9 [39.4, 51.5] µmol/l (no perfusion deficits 52,9 [48.1, 55.1] µmol/l, p = 0.011) and 69,5 [62.4, 74.3] µmol/l (no perfusion deficits 75 [70,80] µmol/l, p = 0.008), respectively. In patients with perfusion deficits, PbtO2 showed no differences in both time frames. PbtO2 was significantly lower in patients with infarctions in both time frames. The median PbtO2 was 17,3 [8,25] mmHg (with no infarctions 29 [22.5, 36] mmHg, p = 0.006) in the 24 h time frame and 21,6 [11.1, 26.4] mmHg (with no infarctions 31 [22,35] mmHg, p = 0.042) in the 48 h time frame. In patients with infarctions, the median values of parameters measured by NIRS showed no significant differences. CONCLUSIONS: The combined NIRS-ICP probe may be useful for early detection of cerebral perfusion deficits and impending DCI. Validation in larger patient collectives is needed.

OxyHbMeter-a novel bedside medical device for monitoring cell-free hemoglobin in the cerebrospinal fluid-proof of principle.

Delayed cerebral ischemia (DCI) occurs in up to one third of patients suffering from aneurysmal subarachnoid hemorrhage (aSAH). Untreated, it leads to secondary cerebral infarctions and is frequently associated with death or severe disability. After aneurysm rupture, erythrocytes in the subarachnoid space lyse and liberate free hemoglobin (Hb), a key driver for the development of DCI. Hemoglobin in the cerebrospinal fluid (CSF-Hb) can be analyzed through a two-step procedure of centrifugation to exclude intact erythrocytes and subsequent spectrophotometric quantification. This analysis can only be done in specialized laboratories but not at the bedside in the intensive care unit. This limits the number of tests done, increases the variability of the results and restricts accuracy. Bedside measurements of CSF-Hb as a biomarker with a point of care diagnostic test system would allow for a continuous monitoring for the risk of DCI in the individual patient. In this study, a microfluidic chip was explored that allows to continuously separate blood particles from CSF or plasma based on acoustophoresis. An in vitro test bench was developed to test in-line measurements with the developed microfluidic chip and a spectrometer. The proof of principle for a continuous particle separation device has been established with diluted blood and CSF samples from animals and aSAH patients, respectively. Processing 1 mL of blood in our microfluidic device was achieved within around 70 min demonstrating only minor deviations from the gold standard centrifugation (7% average error of patient samples), while saving several hours of processing time and additionally the reduction of deviations in the results due to manual labor.

A supervised, externally validated machine learning model for artifact and drainage detection in high-resolution intracranial pressure monitoring data.

OBJECTIVE: In neurocritical care, data from multiple biosensors are continuously measured, but only sporadically acknowledged by the attending physicians. In contrast, machine learning (ML) tools can analyze large amounts of data continuously, taking advantage of underlying information. However, the performance of such ML-based solutions is limited by different factors, for example, by patient motion, manipulation, or, as in the case of external ventricular drains (EVDs), the drainage of CSF to control intracranial pressure (ICP). The authors aimed to develop an ML-based algorithm that automatically classifies normal signals, artifacts, and drainages in high-resolution ICP monitoring data from EVDs, making the data suitable for real-time artifact removal and for future ML applications. METHODS: In their 2-center retrospective cohort study, the authors used labeled ICP data from 40 patients in the first neurocritical care unit (University Hospital Zurich) for model development. The authors created 94 descriptive features that were used to train the model. They compared histogram-based gradient boosting with extremely randomized trees after building pipelines with principal component analysis, hyperparameter optimization via grid search, and sequential feature selection. Performance was measured with nested 5-fold cross-validation and multiclass area under the receiver operating characteristic curve (AUROC). Data from 20 patients in a second, independent neurocritical care unit (Charité - Universitätsmedizin Berlin) were used for external validation with bootstrapping technique and AUROC. RESULTS: In cross-validation, the best-performing model achieved a mean AUROC of 0.945 (95% CI 0.92-0.969) on the development dataset. On the external validation dataset, the model performed with a mean AUROC of 0.928 (95% CI 0.908-0.946) in 100 bootstrapping validation cycles to classify normal signals, artifacts, and drainages. CONCLUSIONS: Here, the authors developed a well-performing supervised model with external validation that can detect normal signals, artifacts, and drainages in ICP signals from patients in neurocritical care units. For future analyses, this is a powerful tool to discard artifacts or to detect drainage events in ICP monitoring signals.

Solving the explainable AI conundrum by bridging clinicians' needs and developers' goals.

Explainable artificial intelligence (XAI) has emerged as a promising solution for addressing the implementation challenges of AI/ML in healthcare. However, little is known about how developers and clinicians interpret XAI and what conflicting goals and requirements they may have. This paper presents the findings of a longitudinal multi-method study involving 112 developers and clinicians co-designing an XAI solution for a clinical decision support system. Our study identifies three key differences between developer and clinician mental models of XAI, including opposing goals (model interpretability vs. clinical plausibility), different sources of truth (data vs. patient), and the role of exploring new vs. exploiting old knowledge. Based on our findings, we propose design solutions that can help address the XAI conundrum in healthcare, including the use of causal inference models, personalized explanations, and ambidexterity between exploration and exploitation mindsets. Our study highlights the importance of considering the perspectives of both developers and clinicians in the design of XAI systems and provides practical recommendations for improving the effectiveness and usability of XAI in healthcare.

Single-antiplatelet regimen in ruptured cerebral blood blister and dissecting aneurysms treated with flow-diverter stent reconstruction.

BACKGROUND: Flow diversion treatment of ruptured cerebral aneurysms remains challenging due to the need for double-antiplatelet therapy. We report our experience with flow-diverter stent (FDS) reconstruction with single-antiplatelet therapy of ruptured cerebral blood blister and dissecting aneurysms. METHODS: In this case series we performed a retrospective analysis of all patients with ruptured cerebral aneurysms who were treated with a phosphoryl-bonded FDS between 2019 and 2022 in a single center. Periprocedurally, all patients received weight-adapted eptifibatide IV and heparin IV. After 6-24 hours, eptifibatide was switched to oral prasugrel as monotherapy. We analyzed the rate of bleeding complications, thromboembolic events, occlusion rate and clinical outcome. RESULTS: Nine patients with subarachnoid hemorrhage were treated, eight within 24 hours of symptom onset. Seven patients were treated with one FDS and two patients received two FDS in a telescopic fashion. Two aneurysms were additionally coil embolized. Fatal re-rupture occurred in one case; eight patients survived and had no adverse events associated with the FDS. Six patients showed complete occlusion of the aneurysm after 3 months (n=2) and 1 year (n=4), respectively. Two patients showed subtotal occlusion of the aneurysm at the last follow-up after 3 months and 6 months, respectively. Favorable clinical outcome was achieved in five patients. CONCLUSIONS: Peri-interventional single-antiplatelet therapy with eptifibatide followed by prasugrel was sufficient to prevent thromboembolic events and reduce re-bleeding using an anti-thrombogenic FDS. FDS with single-antiplatelet therapy might be a viable option for ruptured blood blister and dissecting cerebral aneurysms.

Functional relationship of arterial blood pressure, central venous pressure and intracranial pressure in the early phase after subarachnoid hemorrhage.

BACKGROUND: Intracranial pressure (ICP) and arterial blood pressure (ABP) are related to each other through cerebral autoregulation. Central venous pressure (CVP) is often measured to estimate cardiac filling pressures as an approximate measure for the volume status of a patient. Prior modelling efforts have formalized the functional relationship between CVP, ICP and ABP. However, these models were used to explain short segments of data during controlled experiments and have not yet been used to explain the slowly evolving ICP increase that occurs typically in patients after aneurysmal subarachnoid hemorrhage (SAH). OBJECTIVE: To analyze the functional relationship between ICP, ABP and CVP recorded from SAH patients in the first five days after aneurysm. METHODS: Two methods were used to elucidate this relationship on the running average of the signals: First, using Spearman correlation coefficients calculated over 30 min segments Second, for each patient, linear state space models of ICP as the output and ABP and CVP as inputs were estimated. RESULTS: The mean and variance of the data and the correlation coefficients between ICP-ABP and ICP-CVP vary over time as the patient progresses through their stay in the ICU. On average, after an SAH event, the models show that a) ABP is the bigger driver of changes in ICP than CVP and that increasing ABP leads to reduction in ICP and (b) increasing CVP leads to an increase in ICP. CONCLUSIONS: Finding a) agrees with the hypothesis that patients with subarachnoid hemorrhage have defective autoregulation, and b) agrees with the positive correlation observed between central venous pressure and intracranial pressure in the literature.

Functional NIRS to detect covert consciousness in neurocritical patients.

OBJECTIVE: This pilot study assesses the feasibility to detect covert consciousness in clinically unresponsive patients by means of functional near infrared spectroscopy (fNIRS) in a real intensive care unit setting. We aimed to verify if the hemodynamic response to familiar music measured with fNIRS varies according to the level consciousness of the patients. METHODS: 22 neurocritical patients and 6 healthy controls were included. The experiment consisted in 3 subsequent blocks including a first resting state recording, a period of music playback and a second resting state recording. fNIRS measurement were performed on each subject with two optodes on the forehead. Main oscillatory frequencies of oxyhemoglobin signal were analyzed. Spectral changes of low frequency oscillations (LFO) between subsequent experimental blocks were used as a marker of cortical response. Cortical response was compared to the level of consciousness of the patients and their functional outcome, through validated clinical scores. RESULTS: Cortical hemodynamic response to music on the left prefrontal brain was associated with the level of consciousness of the patients and with their clinical outcome after three months. CONCLUSIONS: Variations in LFO spectral power measured with fNIRS may be a new marker of cortical responsiveness to detect covert consciousness in neurocritical patients. Left prefrontal cortex may play an important role in the perception of familiar music. SIGNIFICANCE: We showed the feasibility of a simple fNIRS approach to detect cortical response in the real setting of an intensive care unit.

ICU Cockpit: a platform for collecting multimodal waveform data, AI-based computational disease modeling and real-time decision support in the intensive care unit.

ICU Cockpit: a secure, fast, and scalable platform for collecting multimodal waveform data, online and historical data visualization, and online validation of algorithms in the intensive care unit. We present a network of software services that continuously stream waveforms from ICU beds to databases and a web-based user interface. Machine learning algorithms process the data streams and send outputs to the user interface. The architecture and capabilities of the platform are described. Since 2016, the platform has processed over 89 billion data points (N = 979 patients) from 200 signals (0.5-500 Hz) and laboratory analyses (once a day). We present an infrastructure-based framework for deploying and validating algorithms for critical care. The ICU Cockpit is a Big Data platform for critical care medicine, especially for multimodal waveform data. Uniquely, it allows algorithms to seamlessly integrate into the live data stream to produce clinical decision support and predictions in clinical practice.