Delirium screening in the intensive care unit using emerging QEEG techniques: A pilot study

A Dual-Camera Eye-Tracking Platform for Rapid Real-Time Diagnosis of Acute Delirium: A Pilot Study

OBJECTIVE

Delirium, an acute confusional state, affects 20-80% of patients in Intensive Care Units (ICUs), one in three medically hospitalized patients, and up to 50% of all patients who have had surgery. Its development is associated with short- and long-term morbidity, and increased risk of death. Yet, we lack any rapid, objective, and automated method to diagnose delirium. Here, we detail the prospective deployment of a novel dual-camera contextual eye-tracking platform. We then use the data from this platform to contemporaneously classify delirium.

RESULTS

We recruited 42 patients, resulting in 210 (114 with delirium, 96 without) recordings of hospitalized patients in ICU across two centers, as part of a prospective multi-center feasibility pilot study. All recordings made with our platform were usable for analysis. We divided the collected data into training and validation cohorts based on the data originating center. We trained two Temporal Convolutional Network (TCN) models that can classify delirium using a pre-existing manual scoring system (Confusion Assessment Method in ICU (CAM-ICU)) as the training target. The first model uses eye movements only which achieves an Area Under the Receiver Operator Curve (AUROC) of 0.67 and a mean Average Precision (mAP) of 0.68. The second model uses the point of regard, the part of the scene the patient is looking at, and increases the AUROC to 0.76 and the mAP to 0.81. These models are the first to classify delirium using continuous non-invasive eye-tracking but will require further clinical prospective validation prior to use as a decision-support tool.

CLINICAL IMPACT

Eye-tracking is a biological signal that can be used to identify delirium in patients in ICU. The platform, alongside the trained neural networks, can automatically, objectively, and continuously classify delirium aiding in the early detection of the deteriorating patient. Future work is aimed at prospective evaluation and clinical translation.

Increased A1 astrocyte activation-driven hippocampal neural network abnormality mediates delirium-like behavior in aged mice undergoing cardiac surgery

Delirium is the most common neurological complication after cardiac surgery with adverse impacts on surgical outcomes. Advanced age is an independent risk factor for delirium occurrence but its underlying mechanisms are not fully understood. Although increased A1 astrocytes and abnormal hippocampal networks are involved in neurodegenerative diseases, whether A1 astrocytes and hippocampal network changes are involved in the delirium-like behavior of aged mice remains unknown. In the present study, a mice model of myocardial ischemia-reperfusion mimicking cardiac surgery and various assessments were used to investigate the different susceptibility of the occurrence of delirium-like behavior between young and aged mice and the underlying mechanisms. The results showed that surgery significantly increased hippocampal A1 astrocyte activation in aged compared to young mice. The high neuroinflammatory state induced by surgery resulted in glutamate accumulation in the extrasynaptic space, which subsequently decreased the excitability of pyramidal neurons and increased the PV interneurons inhibition through enhancing N-methyl-D-aspartate receptors' tonic currents in the hippocampus. These further induced the abnormal activities of the hippocampal neural networks and consequently contributed to delirium-like behavior in aged mice. Notably, the intraperitoneal administration of exendin-4, a glucagon-like peptide-1 receptor agonist, downregulated A1 astrocyte activation and alleviated delirium-like behavior in aged mice, while IL-1α, TNF-α, and C1q in combination administered intracerebroventricularly upregulated A1 astrocyte activation and induced delirium-like behavior in young mice. Therefore, our study suggested that cardiac surgery increased A1 astrocyte activation which subsequently impaired the hippocampal neural networks and triggered delirium development.

Electroencephalogram measured functional connectivity for delirium detection: a systematic review

Objective

Delirium is an acute alteration of consciousness marked by confusion, inattention, and changes in cognition. Some speculate that delirium may be a disorder of functional connectivity, but the requirement to lay still may limit measurement with existing functional imaging modalities in this population. Electroencephalography (EEG) may allow for a more feasible approach to the study of potential connectivity disturbances in delirium. We conducted a systematic review to investigate whether there are EEG-measurable differences in brain functional connectivity in the resting state associated with delirium.

Methods

Medline, PubMed, PsychInfo, Embase and CINAHL were searched for relevant articles containing original data studying EEG functional connectivity measures in delirium.

Results

The search yielded 1,516 records. Following strict inclusion criteria, four studies were included in the review. The studies used a variety of EEG measures including phase lag index, coherence, entropy, shortest path length, minimum spanning tree, and network clustering coefficients to study functional connectivity between scalp electrodes. Across connectivity measures, delirium was associated with decreased brain functional connectivity. All four studies found decreased alpha band connectivity for patients with delirium. None of the studies directly compared the different motor subtypes of delirium.

Significance

This systematic review provides converging evidence for disturbances in oscillatory-based functional connectivity in delirium.

Supervised deep learning with vision transformer predicts delirium using limited lead EEG

As many as 80% of critically ill patients develop delirium increasing the need for institutionalization and higher morbidity and mortality. Clinicians detect less than 40% of delirium when using a validated screening tool. EEG is the criterion standard but is resource intensive thus not feasible for widespread delirium monitoring. This study evaluated the use of limited-lead rapid-response EEG and supervised deep learning methods with vision transformer to predict delirium. This proof-of-concept study used a prospective design to evaluate use of supervised deep learning with vision transformer and a rapid-response EEG device for predicting delirium in mechanically ventilated critically ill older adults. Fifteen different models were analyzed. Using all available data, the vision transformer models provided 99.9%+ training and 97% testing accuracy across models. Vision transformer with rapid-response EEG is capable of predicting delirium. Such monitoring is feasible in critically ill older adults. Therefore, this method has strong potential for improving the accuracy of delirium detection, providing greater opportunity for individualized interventions. Such an approach may shorten hospital length of stay, increase discharge to home, decrease mortality, and reduce the financial burden associated with delirium.

Free-L-Cysteine improves corticosterone-induced behavioral deficits, oxidative stress and neurotransmission in rats

L-Cysteine (L-Cys) is a semi-essential amino acid. It serves as a substrate for enzyme cystathionine-β-synthase in the central nervous system (CNS). L-Cys showed various antioxidant characteristics. Though, studies on the effect of free L-Cys administration to evaluate the CNS functioning is very limited. Therefore, we assessed the effects of L-Cys on corticosterone (CORT) induced oxidative stress, behavioral deficits and memory impairment in male rats. L-Cys (150 mg/kg/ml) administered to vehicle and CORT (20 mg/kg/ml) treated rats orally for 28 days. Behavioral activities were conducted after treatment period. Subsequently, rats were sacrificed, blood and brain were removed. Hippocampus was isolated from brain and then hippocampus and plasma were collected for oxidative, biochemical and neurochemical analysis. Results showed that repeated treatment of L-Cys produced antidepressant, anxiolytic and memory-improving effects which may be ascribed to the enhanced antioxidant profile, normalized cholinergic, serotonergic neurotransmission in brain (hippocampus) following CORT administration. Increased plasma CORT by CORT administration was also normalized by L-Cys. The current study concluded that administration of free L-Cys improved the behavioral, biochemical, neurochemical and redox status of CNS. Hence, L-Cys could be protective therapeutic modulator against stress induced neurological ailments.

Delirium detection using GAMMA wave and machine learning: A pilot study

Delirium occurs in as many as 80% of critically ill older adults and is associated with increased long-term cognitive impairment, institutionalization, and mortality. Less than half of delirium cases are identified using currently available subjective assessment tools. Electroencephalogram (EEG) has been identified as a reliable objective measure but has not been feasible. This study was a prospective pilot proof-of-concept study, to examine the use of machine learning methods evaluating the use of gamma band to predict delirium from EEG data derived from a limited lead rapid response handheld device. Data from 13 critically ill participants aged 50 or older requiring mechanical ventilation for more than 12 h were enrolled. Across the three models, accuracy of predicting delirium was 70 or greater. Stepwise discriminant analysis provided the best overall method. While additional research is needed to determine the best cut points and efficacy, use of a handheld limited lead rapid response EEG device capable of monitoring all five cerebral lobes of the brain for predicting delirium hold promise.

Processed Electroencephalogram-Based Monitoring to Guide Sedation in Critically Ill Adult Patients: Recommendations from an International Expert Panel-Based Consensus

BACKGROUND

The use of processed electroencephalography (pEEG) for depth of sedation (DOS) monitoring is increasing in anesthesia; however, how to use of this type of monitoring for critical care adult patients within the intensive care unit (ICU) remains unclear.

METHODS

A multidisciplinary panel of international experts consisting of 21 clinicians involved in monitoring DOS in ICU patients was carefully selected on the basis of their expertise in neurocritical care and neuroanesthesiology. Panelists were assigned four domains (techniques for electroencephalography [EEG] monitoring, patient selection, use of the EEG monitors, competency, and training the principles of pEEG monitoring) from which a list of questions and statements was created to be addressed. A Delphi method based on iterative approach was used to produce the final statements. Statements were classified as highly appropriate or highly inappropriate (median rating ≥ 8), appropriate (median rating ≥ 7 but < 8), or uncertain (median rating < 7) and with a strong disagreement index (DI) (DI < 0.5) or weak DI (DI ≥ 0.5 but < 1) consensus.

RESULTS

According to the statements evaluated by the panel, frontal pEEG (which includes a continuous colored density spectrogram) has been considered adequate to monitor the level of sedation (strong consensus), and it is recommended by the panel that all sedated patients (paralyzed or nonparalyzed) unfit for clinical evaluation would benefit from DOS monitoring (strong consensus) after a specific training program has been performed by the ICU staff. To cover the gap between knowledge/rational and routine application, some barriers must be broken, including lack of knowledge, validation for prolonged sedation, standardization between monitors based on different EEG analysis algorithms, and economic issues.

CONCLUSIONS

Evidence on using DOS monitors in ICU is still scarce, and further research is required to better define the benefits of using pEEG. This consensus highlights that some critically ill patients may benefit from this type of neuromonitoring.

Using Electroencephalogram Biosignal Changes for Delirium Detection in Intensive Care Units

ABSTRACT

BACKGROUND: Biosignal data acquired during quantitative electroencephalography (QEEG) research may ultimately be used to develop algorithms for more accurate detection of delirium. This study investigates the biosignal changes during delirium states by using the QEEG data of patients in a medical intensive care unit. METHODS: This observational study was conducted between September 2018 and December 2019 at a tertiary hospital in South Korea. Delirium was measured using the Korean version of Confusion Assessment Method for the Intensive Care Unit in intensive care unit patients. Quantitative EEG measurements were recorded for 20 minutes in a natural state without external treatment or stimuli, and QEEG data measured in the centroparietal and parietal regions with eyes open were selected for analysis. Power spectrum analysis with a 5-minute epoch was conducted on the selected 65 cases. RESULTS: QEEG changes in the presence of delirium indicated that alpha, beta, gamma, and spectral edge frequency 50% waves showed significantly lower absolute power spectra than the corresponding findings in the absence of delirium. Brain-mapping results showed that these brain waves were inactivated in delirious states. CONCLUSION: QEEG assessments can potentially detect the changes in the centroparietal and parietal regions of delirium patients. QEEG changes, including lower power spectra of alpha, beta, and gamma waves, and spectral edge frequency 50%, can be successfully used to distinguish delirium from the absence of delirium.

Electroencephalography in delirium assessment: a scoping review

BACKGROUND

Delirium is a common disorder affecting around 31% of patients in the intensive care unit (ICU). Delirium assessment scores such as the Confusion Assessment Method (CAM) are time-consuming, they cannot differentiate between different types of delirium and their etiologies, and they may have low sensitivities in the clinical setting. While today, electroencephalography (EEG) is increasingly being applied to delirious patients in the ICU, a lack of clear cut EEG signs, leads to inconsistent assessments.

METHODS

We therefore conducted a scoping review on EEG findings in delirium. One thousand two hundred thirty-six articles identified through database search on PubMed and Embase were reviewed. Finally, 33 original articles were included in the synthesis.

RESULTS

EEG seems to offer manifold possibilities in diagnosing delirium. All 33 studies showed a certain degree of qualitative or quantitative EEG alterations in delirium. Thus, normal routine (rEEG) and continuous EEG (cEEG) make presence of delirium very unlikely. All 33 studies used different research protocols to at least some extent. These include differences in time points, duration, conditions, and recording methods of EEG, as well as different patient populations, and diagnostic methods for delirium. Thus, a quantitative synthesis and common recommendations are so far elusive.

CONCLUSION

Future studies should compare the different methods of EEG recording and evaluation to identify robust parameters for everyday use. Evidence for quantitative bi-electrode delirium detection based on increased relative delta power and decreased beta power is growing and should be further pursued. Additionally, EEG studies on the evolution of a delirium including patient outcomes are needed.

Electroencephalography of mechanically ventilated patients at high risk of delirium

Objective

Neurophysiological exploration of ICU delirium is limited. Here, we examined EEG characteristics of medical‐surgical critically ill patients with new‐onset altered consciousness state at high risk for ICU delirium.

Materials and methods

Pre‐planned analysis of non‐neurological mechanically ventilated medical‐surgical ICU subjects, who underwent a prospective multicenter randomized, controlled EEG study (NCT03129438, April 2017–November 2018). EEG characteristics, according to the 2012 ACNS nomenclature, included background activity, rhythmic periodic patterns/epileptic activity, amplitude, frequency, stimulus‐induced discharges, triphasic waves, reactivity, and NREM sleep. We explored EEG findings in delirious versus non‐delirious patients, specifically focusing on the presence of burst‐suppression and rhythmic periodic patterns (ictal‐interictal continuum), and ictal activity.

Results

We analyzed 91 patients (median age, 66 years) who underwent EEG because of new‐onset altered consciousness state at a median 5 days from admission; 42 patients developed delirium (46%). Burst‐suppression (10 vs 0%, p = .02), rhythmic/periodic patterns (43% vs 22%, p = .03) and epileptiform activity (7 vs 0%, p = .05) were more frequent in delirious versus non‐delirious patients. The presence of at least one of these abnormal EEG findings (32/91 patients; 35%) was associated with a significant increase in the likelihood of delirium (42 vs 15%, p = .006). Cumulative dose of sedatives and analgesics, as well as all other EEG characteristics, did not differ significantly between the two groups.

Conclusion

In mechanically ventilated non‐neurological critically ill patients with new‐onset alteration of consciousness, EEG showing burst‐suppression, rhythmic or periodic patterns, or seizures/status epilepticus indicate an increased risk of ICU delirium.