A Deep Transfer Learning Framework for Sleep Stage Classification with Single-Channel EEG Signals.

The polysomnogram (PSG) is the gold standard for evaluating sleep quality and disorders. Attempts to automate this process have been hampered by the complexity of the PSG signals and heterogeneity among subjects and recording hardwares. Most of the existing methods for automatic sleep stage scoring rely on hand-engineered features that require prior knowledge of sleep analysis. This paper presents an end-to-end deep transfer learning framework for automatic feature extraction and sleep stage scoring based on a single-channel EEG. The proposed framework was evaluated over the three primary signals recommended by the American Academy of Sleep Medicine (C4-M1, F4-M1, O2-M1) from two data sets that have different properties and are recorded with different hardware. Different Time-Frequency (TF) imaging approaches were evaluated to generate TF representations for the 30 s EEG sleep epochs, eliminating the need for complex EEG signal pre-processing or manual feature extraction. Several training and detection scenarios were investigated using transfer learning of convolutional neural networks (CNN) and combined with recurrent neural networks. Generating TF images from continuous wavelet transform along with a deep transfer architecture composed of a pre-trained GoogLeNet CNN followed by a bidirectional long short-term memory (BiLSTM) network showed the best scoring performance among all tested scenarios. Using 20-fold cross-validation applied on the C4-M1 channel, the proposed framework achieved an average per-class accuracy of 91.2%, sensitivity of 77%, specificity of 94.1%, and precision of 75.9%. Our results demonstrate that without changing the model architecture and the training algorithm, our model could be applied to different single-channel EEGs from different data sets. Most importantly, the proposed system receives a single EEG epoch as an input at a time and produces a single corresponding output label, making it suitable for real time monitoring outside sleep labs as well as to help sleep lab specialists arrive at a more accurate diagnoses.

Data-Driven Prediction for COVID-19 Severity in Hospitalized Patients.

Clinicians urgently need reliable and stable tools to predict the severity of COVID-19 infection for hospitalized patients to enhance the utilization of hospital resources and supplies. Published COVID-19 related guidelines are frequently being updated, which impacts its utilization as a stable go-to resource for informing clinical and operational decision-making processes. In addition, many COVID-19 patient-level severity prediction tools that were developed during the early stages of the pandemic failed to perform well in the hospital setting due to many challenges including data availability, model generalization, and clinical validation. This study describes the experience of a large tertiary hospital system network in the Middle East in developing a real-time severity prediction tool that can assist clinicians in matching patients with appropriate levels of needed care for better management of limited health care resources during COVID-19 surges. It also provides a new perspective for predicting patients' COVID-19 severity levels at the time of hospital admission using comprehensive data collected during the first year of the pandemic in the hospital. Unlike many previous studies for a similar population in the region, this study evaluated 4 machine learning models using a large training data set of 1386 patients collected between March 2020 and April 2021. The study uses comprehensive COVID-19 patient-level clinical data from the hospital electronic medical records (EMR), vital sign monitoring devices, and Polymerase Chain Reaction (PCR) machines. The data were collected, prepared, and leveraged by a panel of clinical and data experts to develop a multi-class data-driven framework to predict severity levels for COVID-19 infections at admission time. Finally, this study provides results from a prospective validation test conducted by clinical experts in the hospital. The proposed prediction framework shows excellent performance in concurrent validation (n=462 patients, March 2020-April 2021) with highest discrimination obtained with the random forest classification model, achieving a macro- and micro-average area under receiver operating characteristics curve (AUC) of 0.83 and 0.87, respectively. The prospective validation conducted by clinical experts (n=185 patients, April-May 2021) showed a promising overall prediction performance with a recall of 78.4-90.0% and a precision of 75.0-97.8% for different severity classes.

COVID-19 Intelligence-Driven Operational Response Platform: Experience of a Large Tertiary Multihospital System in the Middle East.

The COVID-19 pandemic has resulted in global disruptions within healthcare systems, leading to quick dynamic fluctuations in hospital operations and supply chain management. During the early months of the pandemic, tertiary multihospital systems were highly viewed as the go-to hospitals for handling these rapid healthcare challenges caused by the rapidly increasing number of COVID-19 cases. Yet, this pandemic has created an urgent need for coordinated mechanisms to alleviate increasing pressures on these large multihospital systems and ensure services remain high-quality, accessible, and sustainable. Digital health solutions have been identified as promising approaches to address these challenges. This case report describes results for developing multidisciplinary visualizations to support digital health operations in one of the largest tertiary multihospital systems in the Middle East. The report concludes with some lessons and insights learned from the rapid development and delivery of this user-centric COVID-19 multihospital operations intelligent platform.

Deep Recurrent Neural Networks for Automatic Detection of Sleep Apnea from Single Channel Respiration Signals.

Sleep apnea is a common sleep disorder that causes repeated breathing interruption during sleep. The performance of automated apnea detection methods based on respiratory signals depend on the signals considered and feature extraction methods. Moreover, feature engineering techniques are highly dependent on the experts' experience and their prior knowledge about different physiological signals and conditions of the subjects. To overcome these problems, a novel deep recurrent neural network (RNN) framework is developed for automated feature extraction and detection of apnea events from single respiratory channel inputs. Long short-term memory (LSTM) and bidirectional long short-term memory (BiLSTM) are investigated to develop the proposed deep RNN model. The proposed framework is evaluated over three respiration signals: Oronasal thermal airflow (FlowTh), nasal pressure (NPRE), and abdominal respiratory inductance plethysmography (ABD). To demonstrate our results, we use polysomnography (PSG) data of 17 patients with obstructive, central, and mixed apnea events. Our results indicate the effectiveness of the proposed framework in automatic extraction for temporal features and automated detection of apneic events over the different respiratory signals considered in this study. Using a deep BiLSTM-based detection model, the NPRE signal achieved the highest overall detection results with true positive rate (sensitivity) = 90.3%, true negative rate (specificity) = 83.7%, and area under receiver operator characteristic curve = 92.4%. The present results contribute a new deep learning approach for automated detection of sleep apnea events from single channel respiration signals that can potentially serve as a helpful and alternative tool for the traditional PSG method.

An Improved Passivity-based Control of Electrostatic MEMS Device.

It is commonly known that the performance of an electrostatic microelectromechanical system (MEMS) device is limited to a specific range of the full gap distance due to the inherited "pull-in instability" phenomenon. In this work, we design a controller to extend the stabilization range of an electrostatic MEMS device and to enhance its performance. The interconnection and damping assignment-passivity based control (IDA-PBC) method is used and the controller design involves coordinate transformations and a coupling between the mechanical and electrical subsystems of the device. The method deploys a design of a speed observer to estimate the speed, which cannot be directly measured by sensors. The effectiveness of the dynamical controller is verified via numerical simulations; it is evident by the extended travel range of the parallel plates as well as the improved performance of the plates, even with a naturally lighter damping ratio.

Short-Time Wind Speed Forecast Using Artificial Learning-Based Algorithms.

The need for an efficient power source for operating the modern industry has been rapidly increasing in the past years. Therefore, the latest renewable power sources are difficult to be predicted. The generated power is highly dependent on fluctuated factors (such as wind bearing, pressure, wind speed, and humidity of surrounding atmosphere). Thus, accurate forecasting methods are of paramount importance to be developed and employed in practice. In this paper, a case study of a wind harvesting farm is investigated in terms of wind speed collected data. For data like the wind speed that are hard to be predicted, a well built and tested forecasting algorithm must be provided. To accomplish this goal, four neural network-based algorithms: artificial neural network (ANN), convolutional neural network (CNN), long short-term memory (LSTM), and a hybrid model convolutional LSTM (ConvLSTM) that combines LSTM with CNN, and one support vector machine (SVM) model are investigated, evaluated, and compared using different statistical and time indicators to assure that the final model meets the goal that is built for. Results show that even though SVM delivered the most accurate predictions, ConvLSTM was chosen due to its less computational efforts as well as high prediction accuracy.

Time domain characterization for sleep apnea in oronasal airflow signal: a dynamic threshold classification approach.

OBJECTIVE: Apneas are the most common type of sleep-related breathing disorders; they cause patients to move from restorative sleep into inefficient sleep. The American Academy of Sleep Medicine (AASM) considers sleep apnea as a hidden health crisis that affects 29.4 million adults, costing the USA billions of dollars. Traditional detection methods of sleep apnea are achieved by human observation of the respiration signals. This introduces limitations in terms of access and efficiency of diagnostic sleep studies. However, alternative device technologies have limited diagnostic accuracy for detecting apnea events although many of the previous investigational algorithms are based on multiple physiological channel inputs. Guided by the AASM recommendations for sleep apnea diagnostics, this paper investigates time domain metrics to characterize changes in oronasal airflow respiration signals during the occurrence of apneic events. APPROACH: A new algorithm is developed to derive a respiratory baseline from the oronasal airflow signal in order to detect sleep apnea events using a dynamically adjusted threshold classification approach. To demonstrate our results, we use polysomnography data of [Formula: see text] patients with different apnea severity levels as reflected by their overnight apnea hypopnea index (AHI), including patients with mild apnea (5 [Formula: see text] AHI [Formula: see text]), moderate apnea ([Formula: see text] AHI [Formula: see text]), and severe apnea (AHI [Formula: see text]). MAIN RESULTS: Our results indicate the ability to characterize sleep apnea events in oronasal airflow signals using the proposed dynamic threshold classification approach. Overall, the new algorithm achieved a sensitivity of 80.0%, specificity of 88.7%, and an area under receiver operating characteristics curve of 0.844. SIGNIFICANCE: The present results contribute a new approach for progressive detection of sleep apnea using an adaptive threshold that is dynamically adjusted with respect to the patient's respiration baseline, making it potentially able to effectively generalize over patients with different apnea severity levels and longer monitoring periods.

Multi-Step Ahead Predictions for Critical Levels in Physiological Time Series.

Standard modeling and evaluation methods have been classically used in analyzing engineering dynamical systems where the fundamental problem is to minimize the (mean) error between the real and predicted systems. Although these methods have been applied to multi-step ahead predictions of physiological signals, it is often more important to predict clinically relevant events than just to match these signals. Adverse clinical events, which occur after a physiological signal breaches a clinically defined critical threshold, are a popular class of such events. This paper presents a framework for multi-step ahead predictions of critical levels of abnormality in physiological signals. First, a performance metric is presented for evaluating multi-step ahead predictions. Then, this metric is used to identify personalized models optimized with respect to predictions of critical levels of abnormality. To address the paucity of adverse events, weighted support vector machines and cost-sensitive learning are used to optimize the proposed framework with respect to statistical metrics that can take into account the relative rarity of such events.

A new algorithm for the detection of sleep apnea events in respiration signals.

Sleep apneas are the most common type of sleep-related breathing disorders which cause a patient to move from a good sleep into an inefficient sleep. In addition, sleep apnea widely impacts the American population and is a large cost for healthcare. Traditional detection methods of sleep apneas are complex, expensive, and invasive to most patients. Among the various physiological signals, respiration signals are relatively easy to be monitored. However, not many studies are conducted using respiration signal only, and most of the previous algorithms are insufficient to detect apnea events. In this paper, we propose a new algorithm based on only the respiration signal to detect the apnea events during sleep and conduct experiments comparing the performance of our algorithm against two apnea detection algorithms. We use 20 patients' data, all of whom have severe Apnea Hypopnea Index (AHI>30: over 30 events per hour). Our study shows that our algorithm outperforms the other two algorithms.

Effect of concurrent oxygen therapy on accuracy of forecasting imminent postoperative desaturation.

Episodic postoperative desaturation occurs predominantly from respiratory depression or airway obstruction. Monitor display of desaturation is typically delayed by over 30 s after these dynamic inciting events, due to perfusion delays, signal capture and averaging. Prediction of imminent critical desaturation could aid development of dynamic high-fidelity response systems that reduce or prevent the inciting event from occurring. Oxygen therapy is known to influence the depth and duration of desaturation epochs, thereby potentially influencing the accuracy of forecasting of desaturation. In this study, postoperative pulse oximetry data were retrospectively modeled using autoregressive methods to create prediction models for [Formula: see text] and imminent critical desaturation in the postoperative period. The accuracy of these models in predicting near future [Formula: see text] values was tested using root mean square error. The model accuracy for prediction of critical desaturation ([Formula: see text] [Formula: see text]) was evaluated using meta-analytical methods (sensitivity, specificity, likelihood ratios, diagnostic odds ratios and area under summary receiver operating characteristic curves). Between-study heterogeneity was used as a measure of reliability of the model across different patients and evaluated using the tau-squared statistic. Model performance was evaluated in [Formula: see text] patients who received postoperative oxygen supplementation and [Formula: see text] patients who did not receive oxygen. Our results show that model accuracy was high with root mean square errors between 0.2 and 2.8%. Prediction accuracy as defined by area under the curve for critical desaturation events was observed to be greater in patients receiving oxygen in the 60-s horizon ([Formula: see text] vs. [Formula: see text]). This was likely related to the higher frequency of events in this group (median [IQR] [Formula: see text] [Formula: see text]) than patients who were not treated with oxygen ([Formula: see text] [Formula: see text]; [Formula: see text]). Model reliability was reflected by the homogeneity of the prediction models which were homogenous across both prediction horizons and oxygen treatment groups. In conclusion, we report the use of autoregressive models to predict [Formula: see text] and forecast imminent critical desaturation events in the postoperative period with high degree of accuracy. These models reliably predict critical desaturation in patients receiving supplemental oxygen therapy. While high-fidelity prophylactic interventions that could modify these inciting events are in development, our current study offers proof of concept that the afferent limb of such a system can be modeled with a high degree of accuracy.

Evaluating predictions of critical oxygen desaturation events.

This paper presents a new approach for evaluating predictions of oxygen saturation levels in blood ( SpO2). A performance metric based on a threshold is proposed to evaluate SpO2 predictions based on whether or not they are able to capture critical desaturations in the SpO2 time series of patients. We use linear auto-regressive models built using historical SpO2 data to predict critical desaturation events with the proposed metric. In 20 s prediction intervals, 88%-94% of the critical events were captured with positive predictive values (PPVs) between 90% and 99%. Increasing the prediction horizon to 60 s, 46%-71% of the critical events were detected with PPVs between 81% and 97%. In both prediction horizons, more than 97% of the non-critical events were correctly classified. The overall classification capabilities for the developed predictive models were also investigated. The area under ROC curves for 60 s predictions from the developed models are between 0.86 and 0.98. Furthermore, we investigate the effect of including pulse rate (PR) dynamics in the models and predictions. We show no improvement in the percentage of the predicted critical desaturations if PR dynamics are incorporated into the SpO2 predictive models (p-value = 0.814). We also show that including the PR dynamics does not improve the earliest time at which critical SpO2 levels are predicted (p-value = 0.986). Our results indicate oxygen in blood is an effective input to the PR rather than vice versa. We demonstrate that the combination of predictive models with frequent pulse oximetry measurements can be used as a warning of critical oxygen desaturations that may have adverse effects on the health of patients.