Using artificial intelligence to predict adverse outcomes in emergency department patients with hyperglycemic crises in real time.

BACKGROUND: Hyperglycemic crises are associated with high morbidity and mortality. Previous studies have proposed methods to predict adverse outcomes of patients in hyperglycemic crises; however, artificial intelligence (AI) has never been used to predict adverse outcomes. We implemented an AI model integrated with the hospital information system (HIS) to clarify whether AI could predict adverse outcomes. METHODS: We included 2,666 patients with hyperglycemic crises from emergency departments (ED) between 2009 and 2018. The patients were randomized into a 70%/30% split for AI model training and testing. Twenty-two feature variables from the electronic medical records were collected. The performance of the multilayer perceptron (MLP), logistic regression, random forest, Light Gradient Boosting Machine (LightGBM), support vector machine (SVM), and K-nearest neighbor (KNN) algorithms was compared. We selected the best algorithm to construct an AI model to predict sepsis or septic shock, intensive care unit (ICU) admission, and all-cause mortality within 1 month. The outcomes between the non-AI and AI groups were compared after implementing the HIS and predicting the hyperglycemic crisis death (PHD) score. RESULTS: The MLP had the best performance in predicting the three adverse outcomes, compared with the random forest, logistic regression, SVM, KNN, and LightGBM models. The areas under the curves (AUCs) using the MLP model were 0.852 for sepsis or septic shock, 0.743 for ICU admission, and 0.796 for all-cause mortality. Furthermore, we integrated the AI predictive model with the HIS to assist decision making in real time. No significant differences in ICU admission or all-cause mortality were detected between the non-AI and AI groups. The AI model performed better than the PHD score for predicting all-cause mortality (AUC 0.796 vs. 0.693). CONCLUSIONS: A real-time AI predictive model is a promising method for predicting adverse outcomes in ED patients with hyperglycemic crises. Further studies recruiting more patients are warranted.

Real-time interactive artificial intelligence of things-based prediction for adverse outcomes in adult patients with pneumonia in the emergency department.

BACKGROUND: Artificial intelligence of things (AIoT) may be a solution for predicting adverse outcomes in emergency department (ED) patients with pneumonia; however, this issue remains unclear. Therefore, we conducted this study to clarify it. METHODS: We identified 52,626 adult ED patients with pneumonia from three hospitals between 2010 and 2019 for this study. Thirty-three feature variables from electronic medical records were used to construct an artificial intelligence (AI) model to predict sepsis or septic shock, respiratory failure, and mortality. After comparisons of the predictive accuracies among logistic regression, random forest, support-vector machine (SVM), light gradient boosting machine (LightGBM), multilayer perceptron (MLP), and eXtreme Gradient Boosting (XGBoost), we selected the best one to build the model. We further combined the AI model with the Internet of things as AIoT, added an interactive mode, and implemented it in the hospital information system to assist clinicians with decision making in real time. We also compared the AIoT-based model with the confusion-urea-respiratory rate-blood pressure-65 (CURB-65) and pneumonia severity index (PSI) for predicting mortality. RESULTS: The best AI algorithms were random forest for sepsis or septic shock (area under the curve [AUC] = 0.781), LightGBM for respiratory failure (AUC = 0.847), and mortality (AUC = 0.835). The AIoT-based model represented better performance than CURB-65 and PSI indicators for predicting mortality (0.835 vs. 0.681 and 0.835 vs. 0.728). CONCLUSIONS: A real-time interactive AIoT-based model might be a better tool for predicting adverse outcomes in ED patients with pneumonia. Further validation in other populations is warranted.

Predicting outcomes in older ED patients with influenza in real time using a big data-driven and machine learning approach to the hospital information system.

BACKGROUND: Predicting outcomes in older patients with influenza in the emergency department (ED) by machine learning (ML) has never been implemented. Therefore, we conducted this study to clarify the clinical utility of implementing ML. METHODS: We recruited 5508 older ED patients (≥65 years old) in three hospitals between 2009 and 2018. Patients were randomized into a 70%/30% split for model training and testing. Using 10 clinical variables from their electronic health records, a prediction model using the synthetic minority oversampling technique preprocessing algorithm was constructed to predict five outcomes. RESULTS: The best areas under the curves of predicting outcomes were: random forest model for hospitalization (0.840), pneumonia (0.765), and sepsis or septic shock (0.857), XGBoost for intensive care unit admission (0.902), and logistic regression for in-hospital mortality (0.889) in the testing data. The predictive model was further applied in the hospital information system to assist physicians' decisions in real time. CONCLUSIONS: ML is a promising way to assist physicians in predicting outcomes in older ED patients with influenza in real time. Evaluations of the effectiveness and impact are needed in the future.

Real-time AI prediction for major adverse cardiac events in emergency department patients with chest pain.

BACKGROUND: A big-data-driven and artificial intelligence (AI) with machine learning (ML) approach has never been integrated with the hospital information system (HIS) for predicting major adverse cardiac events (MACE) in patients with chest pain in the emergency department (ED). Therefore, we conducted the present study to clarify it. METHODS: In total, 85,254 ED patients with chest pain in three hospitals between 2009 and 2018 were identified. We randomized the patients into a 70%/30% split for ML model training and testing. We used 14 clinical variables from their electronic health records to construct a random forest model with the synthetic minority oversampling technique preprocessing algorithm to predict acute myocardial infarction (AMI) < 1 month and all-cause mortality < 1 month. Comparisons of the predictive accuracies among random forest, logistic regression, support-vector clustering (SVC), and K-nearest neighbor (KNN) models were also performed. RESULTS: Predicting MACE using the random forest model produced areas under the curves (AUC) of 0.915 for AMI < 1 month and 0.999 for all-cause mortality < 1 month. The random forest model had better predictive accuracy than logistic regression, SVC, and KNN. We further integrated the AI prediction model with the HIS to assist physicians with decision-making in real time. Validation of the AI prediction model by new patients showed AUCs of 0.907 for AMI < 1 month and 0.888 for all-cause mortality < 1 month. CONCLUSIONS: An AI real-time prediction model is a promising method for assisting physicians in predicting MACE in ED patients with chest pain. Further studies to evaluate the impact on clinical practice are warranted.

Orbitofrontal neurons signal reward predictions, not reward prediction errors.

Neurons in the orbitofrontal cortex (OFC) fire in anticipation of and during rewards. Such firing has been suggested to encode reward predictions and to account in some way for the role of this area in adaptive behavior and learning. However, it has also been reported that neural activity in OFC reflects reward prediction errors, which might drive learning directly. Here we tested this question by analyzing the firing of OFC neurons recorded in an odor discrimination task in which rats were trained to sample odor cues and respond left or right on each trial for reward. Neurons were recorded across blocks of trials in which we switched either the number or the flavor of the reward delivered in each well. Previously we have described how neurons in this dataset fired to the predictive cues (Stalnaker et al., 2014); here we focused on the firing in anticipation of and just after delivery of each drop of reward, looking specifically for differences in firing based on whether the reward number or flavor was unexpected or expected. Unlike dopamine neurons recorded in this setting, which exhibited phasic error-like responses after surprising changes in either reward number or reward flavor (Takahashi et al., 2017), OFC neurons showed no such error correlates and instead fired in a way that reflected reward predictions.

Orbitofrontal lesions eliminate signalling of biological significance in cue-responsive ventral striatal neurons.

The ventral striatum has long been proposed as an integrator of biologically significant associative information to drive actions. Although inputs from the amygdala and hippocampus have been much studied, the role of prominent inputs from orbitofrontal cortex (OFC) are less well understood. Here, we recorded single-unit activity from ventral striatum core in rats with sham or ipsilateral neurotoxic lesions of lateral OFC, as they performed an odour-guided spatial choice task. Consistent with prior reports, we found that spiking activity recorded in sham rats during cue sampling was related to both reward magnitude and reward identity, with higher firing rates observed for cues that predicted more reward. Lesioned rats also showed differential activity to the cues, but this activity was unbiased towards larger rewards. These data support a role for OFC in shaping activity in the ventral striatum to represent the biological significance of associative information in the environment.

Orbitofrontal neurons acquire responses to 'valueless' Pavlovian cues during unblocking.

The orbitofrontal cortex (OFC) has been described as signaling outcome expectancies or value. Evidence for the latter comes from the studies showing that neural signals in the OFC correlate with value across features. Yet features can co-vary with value, and individual units may participate in multiple ensembles coding different features. Here we used unblocking to test whether OFC neurons would respond to a predictive cue signaling a 'valueless' change in outcome flavor. Neurons were recorded as the rats learned about cues that signaled either an increase in reward number or a valueless change in flavor. We found that OFC neurons acquired responses to both predictive cues. This activity exceeded that exhibited to a 'blocked' cue and was correlated with activity to the actual outcome. These results show that OFC neurons fire to cues with no value independent of what can be inferred through features of the predicted outcome.

Orbitofrontal neurons infer the value and identity of predicted outcomes.

The best way to respond flexibly to changes in the environment is to anticipate them. Such anticipation often benefits us if we can infer that a change has occurred, before we have actually experienced the effects of that change. Here we test for neural correlates of this process by recording single-unit activity in the orbitofrontal cortex in rats performing a choice task in which the available rewards changed across blocks of trials. Consistent with the proposal that orbitofrontal cortex signals inferred information, firing changes at the start of each new block as if predicting the not-yet-experienced reward. This change occurs whether the new reward is different in number of drops, requiring signalling of a new value, or in flavour, requiring signalling of a new sensory feature. These results show that orbitofrontal neurons provide a behaviourally relevant signal that reflects inferences about both value-relevant and value-neutral information about impending outcomes.

Effects of dopamine D2 agonist quinpirole on neuronal activity of anterior cingulate cortex and striatum in rats.

RATIONALE: The influence of acute D2 agonist quinpirole on locomotor activity has been effectively characterized. However, few studies have addressed the dynamic changes in neuronal activity of the anterior cingulate cortex (ACC) and striatum (STR), two crucial regions for cognitive and motor functions, after quinpirole administration. OBJECTIVE: This study was conducted in order to acquire detailed information on the evoked activity of the neurons in the ACC and STR after acute quinpirole administration. METHODS: Multichannel electrophysiological recording was used for tracking neuronal activity in the ACC and STR of urethane-anesthetized rats after administration of saline or 0.05 or 0.5 mg/kg quinpirole. RESULTS: In contrast to the responses to saline, quinpirole dose-dependently increased the ratio of neurons, the activity of which was inhibited in the ACC and STR. By examining the ensemble neuronal activities of inhibition-responded neurons, there was no significant activity difference among the "treatments" (saline and low- and high-dose quinpirole), the "periods" (the duration of 0-15 and 16-45 min after i.v. injection), and the interaction between "treatments" and "periods." Regarding activation-responded neurons, however, there was a significant "periods" difference in both ACC and STR, and the activity of 16-45 min was significantly higher than the activity of 0-15 min after high-dose quinpirole administration in ACC (p < 0.05) and STR (p < 0.001). CONCLUSION: Dose-dependent ACC and STR neuronal responses to quinpirole may offer a possible mechanism for understanding the locomotor responses to quinpirole in behaving rats. The late excitatory effect of high-dose quinpirole in the STR further suggests that this region would be critical for the activation of locomotor activity.

Does the orbitofrontal cortex signal value?

The orbitofrontal cortex (OFC) has long been implicated in associative learning. Early work by Mishkin and Rolls showed that the OFC was critical for rapid changes in learned behavior, a role that was reflected in the encoding of associative information by orbitofrontal neurons. Over the years, new data-particularly neurophysiological data-have increasingly emphasized the OFC in signaling actual value. These signals have been reported to vary according to internal preferences and judgments and to even be completely independent of the sensory qualities of predictive cues, the actual rewards, and the responses required to obtain them. At the same time, increasingly sophisticated behavioral studies have shown that the OFC is often unnecessary for simple value-based behavior and instead seems critical when information about specific outcomes must be used to guide behavior and learning. Here, we review these data and suggest a theory that potentially reconciles these two ideas, value versus specific outcomes, and bodies of work on the OFC.

Post-training infusion of glutamate into the bed nucleus of the stria terminalis enhanced inhibitory avoidance memory: an effect involving norepinephrine.

This study examined an interaction between glutamate and norepinephrine in the bed nucleus of the stria terminalis (BNST) in modulating affective memory formation. Male Wistar rats with indwelling cannulae in the BNST were trained on a one-trial step-through inhibitory avoidance task and received pre- or post-training intra-BNST infusion of glutamate, norepinephrine or their antagonists. Results of the 1-day test indicated that post-training intra-BNST infusion of DL-2-amino-5-phosphonovaleric acid (APV) impaired retention in a dose- and time-dependent manner, while infusion of glutamate had an opposite effect. Co-infusion of 0.2microg glutamate and 0.02microg norepinephrine resulted in marked retention enhancement by summating non-apparent effects of the two drugs given at a sub-enhancing dose. The amnesic effect of 5.0microg APV was ameliorated by 0.02microg norepinephrine, while the memory enhancing effect of 1.0microg glutamate was attenuated by 5.0microg propranolol. These findings suggest that training on an inhibitory avoidance task may alter glutamate neurotransmission, which by activating NMDA receptors releases norepinephrine to modulate memory formation via beta adrenoceptors in the BNST.

Posttraining infusion of cholinergic drugs into the ventral subiculum modulated memory in an inhibitory avoidance task: interaction with the bed nucleus of the stria terminalis.

The ventral subiculum (vSUB), a hippocampal efferent target implicated in learning and stress coping, receives cholinergic input and sends glutamatergic output to the bed nucleus of the stria terminalis (BNST). This study examined the roles of vSUB muscarinic activation and its interaction with BNST N-methyl-D-aspartate and noradrenergic receptors in formation of aversive memory. Male Wistar rats with cannulae implanted into the vSUB or BNST were trained on a step-through inhibitory avoidance task. Shortly after training, they received cholinergic drugs infused into the vSUB and/or glutamatergic or noradrenergic drugs infused into the BNST. Results of the 1-day retention tests showed that intra-vSUB infusion of oxotremorine (0.01 microg) or scopolamine (0.3 or 3.0 microg) enhanced or impaired retention, respectively. Both effects were dose- and time-dependent, and 0.001 microg oxotremorine attenuated the amnesia induced by 3.0 microg scopolamine. The oxotremorine-induced memory enhancement was blocked by intra-BNST infusion of DL-2-amino-5-phosphonovaleric acid or propranolol at a dose not affecting retention; the amnesia induced by scopolamine was blunted by intra-BNST infusion of glutamate or norepinephrine at a dose with a negligible effect on retention. These data suggest that in an inhibitory avoidance task muscarinic activation of the vSUB modulated memory formation by interacting with the BNST glutamatergic and noradrenergic functions.