Infants are at risk for potentially life-threatening postoperative apnea (POA). We developed an Automated Unsupervised Respiratory Event Analysis (AUREA) to classify breathing patterns obtained with dual belt respiratory inductance plethysmography and a reference using Expectation Maximization (EM). This work describes AUREA and evaluates its performance. AUREA computes six metrics and inputs them into a series of four binary k-means classifiers. Breathing patterns were characterized by normalized variance, nonperiodic power, instantaneous frequency and phase. Signals were classified sample by sample into one of 5 patterns: pause (PAU), movement (MVT), synchronous (SYB) and asynchronous (ASB) breathing, and unknown (UNK). MVT and UNK were combined as UNKNOWN. Twenty-one preprocessed records obtained from infants at risk for POA were analyzed. Performance was evaluated with a confusion matrix, overall accuracy, and pattern specific precision, recall, and F-score. Segments of identical patterns were evaluated for fragmentation and pattern matching with the EM reference. PAU exhibited very low normalized variance. MVT had high normalized nonperiodic power and low frequency. SYB and ASB had a median frequency of respectively, 0.76Hz and 0.71Hz, and a mode for phase of 4o and 100o. Overall accuracy was 0.80. AUREA confused patterns most often with UNKNOWN (25.5%). The pattern specific F-score was highest for SYB (0.88) and lowest for PAU (0.60). PAU had high precision (0.78) and low recall (0.49). Fragmentation was evident in pattern events <2s. In 75% of the EM pattern events >2s, 50% of the samples classified by AUREA had identical patterns. Frequency and phase for SYB and ASB were consistent with published values for synchronous and asynchronous breathing in infants. The low normalized variance in PAU, was consistent with published scoring rules for pediatric apnea. These findings support the use of AUREA to classify breathing patterns and warrant a future evaluation of clinically relevant respiratory events.

Manual scoring (MS) of cardiorespiratory signals is the Manual scoring (MS) of cardiorespiratory signals is the gold standard method for the analysis of respiratory data in sleep laboratories. In MS, trained, expert scorers characterize respiratory patterns by scrolling through a data record and visually identifying patterns. However, MS is limited by high intra- and inter-scorer variability and subjectivity. A strategy to mitigate this is to analyze the same respiratory data multiple times and generate a consensus. This consensus is generally determined by a majority vote (MV), where the most frequent pattern is selected as the true pattern. This paper presents expectation-maximization pattern sequence (EM-PSEQ), a novel method based on EM that estimates the true patterns optimally. A simulation study examined the accuracies of EM-PSEQ, MV, and individual scorers (IS) as a function of the number of analyses. Accuracy was measured with the Fleiss kappa statistic, and is reported as (kappa(MDN) = x; kappa(P5) where kappa(MDN) the median value, is the expected accuracy, and kappa(P5), the 5th percentile value, gives the minimum accuracy for 95% confidence. IS accuracy remained constant at [kappa(MDN) = 0.67; kappa(P5 )= 0.60] as the number of analyses increased. MV accuracy increased slowly with the number of analyses and plateaued at [kappa(MDN) = 0.78: kappa(P5) = 0.761 after five analyses. In contrast, EM-PSEQ accuracy improved quickly, reaching an almost perfect value of (kappa(MDN )= 0.83; kappa(P5) = 0.771 with four analyses, and perfect accuracy [kappa(MDN) = 1.00; kappa(P5 )= 0.991 after 25 analyses. EM-PSEQ performed much better than either MV or IS, and required only modest computational effort. Consequently, we believe EM-PSEQ will be a very valuable tool for clinical studies, as it can dramatically improve the accuracy of manual respiratory analysis with minimal additional cost.

Extremely preterm infants (≤ 28 weeks gestation) commonly require endotracheal intubation and mechanical ventilation (MV) to maintain adequate oxygenation and gas exchange. Given that MV is independently associated with important adverse outcomes, efforts should be made to limit its duration. However, current methods for determining extubation readiness are inaccurate and a significant number of infants fail extubation and require reintubation, an intervention that may be associated with increased morbidities. A variety of objective measures have been proposed to better define the optimal time for extubation, but none have proven clinically useful. In a pilot study, investigators from this group have shown promising results from sophisticated, automated analyses of cardiorespiratory signals as a predictor of extubation readiness. The aim of this study is to develop an automated predictor of extubation readiness using a combination of clinical tools along with novel and automated measures of cardiorespiratory behavior, to assist clinicians in determining when extremely preterm infants are ready for extubation.

BackgroundRates of extubation failure of extremely preterm infants remain high. Analysis of breathing patterns variability during spontaneous breathing under endotracheal tube continuous positive airway pressure (ETT-CPAP) is a potential tool to predict extubation readiness.ObjectiveTo investigate if automated analysis of respiratory signals would reveal differences in respiratory behavior between infants that were successfully extubated or not.MethodsRespiratory Inductive Plethysmography (RIP) signals were recorded during ETT-CPAP just prior to extubation. Signals were digitized, and analyzed using an Automated Unsupervised Respiratory Event Analysis (AUREA). Extubation failure was defined as reintubation within 72 hr. Statistical differences between infants who were successfully extubated or failed were calculated.ResultsA total of 56 infants were enrolled and one was excluded due to instability during the ETT-CPAP; 11 out of 55 infants studied failed extubation (20%). No differences in demographics were observed between the success and failure groups. Significant differences on the variability of some respiratory parameters or metrics' estimated by AUREA were observed between the 2 groups. Indeed, a simple classification using the variability of two metrics of respiratory behavior predicted extubation failure with high accuracy.ConclusionAutomated analysis of respiratory behavior during a short ETT-CPAP period may help in the prediction of extubation readiness in extremely preterm infants. Pediatr Pulmonol. 2015; 50:479-486. (c) 2015 The Authors. Pediatric Pulmonology published by Wiley Periodicals, Inc.

Infants recovering from anesthesia are at risk of life threatening Postoperative Apnea (POA). POA events are rare, and so the study of POA requires the analysis of long cardiorespiratory records. Manual scoring is the preferred method of analysis for these data, but it is limited by low intra- and inter-scorer repeatability. Furthermore, recommended scoring rules do not provide a comprehensive description of the respiratory patterns. This work describes a set of manual scoring tools that address these limitations. These tools include: (i) a set of definitions and scoring rules for 6 mutually exclusive, unique patterns that fully characterize infant respiratory inductive plethysmography (RIP) signals; (ii) RIPScore, a graphical, manual scoring software to apply these rules to infant data; (iii) a library of data segments representing each of the 6 patterns; (iv) a fully automated, interactive formal training protocol to standardize the analysis and establish intra- and inter-scorer repeatability; and (v) a quality control method to monitor scorer ongoing performance over time. To evaluate these tools, three scorers from varied backgrounds were recruited and trained to reach a performance level similar to that of an expert. These scorers used RIPScore to analyze data from infants at risk of POA in two separate, independent instances. Scorers performed with high accuracy and consistency, analyzed data efficiently, had very good intra- and inter-scorer repeatability, and exhibited only minor confusion between patterns. These results indicate that our tools represent an excellent method for the analysis of respiratory patterns in long data records. Although the tools were developed for the study of POA, their use extends to any study of respiratory patterns using RIP (e.g., sleep apnea, extubation readiness). Moreover, by establishing and monitoring scorer repeatability, our tools enable the analysis of large data sets by multiple scorers, which is essential for longitudinal and multicenter studies.

This paper describes organizational guidelines and an anonymization protocol for the management of sensitive information in interdisciplinary, multi-institutional studies with multiple collaborators. This protocol is flexible, automated, and suitable for use in cloud-based projects as well as for publication of supplementary information in journal papers. A sample implementation of the anonymization protocol is illustrated for an ongoing study dealing with Automated Prediction of EXtubation readiness (APEX).

Extremely preterm infants (gestational age ≤ 28 weeks) often require EndoTracheal Tube-Invasive Mechanical Ventilation (ETT-IMV) to survive. Clinicians wean infants off ETT-IMV as early as possible using their judgment and clinical information. However, assessment of extubation readiness is not accurate since 20 to 40% of preterm infants fail extubation. We extended our work in automated prediction of extubation readiness by examining correlations of automated cardiorespiratory features to clinical parameters in successfully extubated infants. Only a few features, mainly those related to variability of breathing synchrony, had any consistent correlation with clinical parameters, namely gestational age, day of life at extubation, and bicarbonate. We conclude that the automated cardiorespiratory features likely provide different information additional to clinical practice.

Infants recovering from general anesthesia are at risk of postoperative apnea (POA), a potentially life threatening event. There is no accurate way to identify which infants will experience POA, and thus all infants with postmenstrual age <; 60 weeks are monitored for apnea in hospital postoperatively. Using a comprehensive, automated analysis of the postoperative breathing patterns, we identified the occurrence of respiratory pauses in 24 infants at age risk for POA. We determined the POA category for each infant by using K-medoids to cluster the duration of the longest respiratory pause. Two clusters were identified, corresponding to APNEA and NO-APNEA, with a threshold of 14.6 s, a value consistent with the clinically accepted threshold of 15 s. K-medoids derived POA labels were used to evaluate the predictive ability of demographic and anesthetic management variables. Weight and the intraoperative doses of atropine, propofol, and opioids discriminated between the APNEA and NO-APNEA groups. A linear Gaussian discriminant analysis classifier provided a very good classification with a probability of detection P D = 0.73 and a probability of false alarm P FA = 0.22. This approach provides a promising tool for the systematic, objective study of infants at risk of POA.

Oximeters are commonly used in abbreviated cardiorespiratory studies (ACS) to monitor blood oxygen saturation and heart rate using the photoplethysmography (PPG) signal. These data are prone to movement artifacts, especially in infants who move or need to be handled often. Therefore segments of PPG data contaminated by movement artifact must be detected as a first stage of analysis. In ACS this identification is generally done manually, by having an expert visually assess the quality of the signal. This is subjective and very time consuming, especially for long data records. For this reason we present a novel detector of PPG movement artifacts that uses moving average filters to remove trends, reduce the effect of white noise, and notch filter pulse-related information. The normalized root mean square of the filtered signal is then used as a detection statistic. We demonstrate its detection properties using a data set from infants recovering from anesthesia, and show that it performs better than other automated methods based on entropy or higher-order statistics. Furthermore, the new method is more robust than the other methods in the presence of large noise.

The typical approach for analysis of respiratory records consists of detection of respiratory pauses and elimination of segments corrupted by movement artifacts. This is motivated by established rules used for manual scoring of respiratory events, which focus on pause segmentation and do not define criteria to identify breathing segments. With this strategy, breathing segments can only be inferred indirectly from the absence of abnormalities, yielding an unclear and ambiguous definition. In this work we present novel detectors for synchronous and asynchronous breathing, and compare them with AUREA, a novel system for Automated Unsupervised Respiratory Event Analysis, which performs indirect classification of breathing. Results from analysis of real infant respiratory data show an improvement in the identification of synchronous and asynchronous breathing of 9% and 27% respectively, demonstrating that direct detection of breathing enhances the classification performance.

The majority of extreme preterm infants require endotracheal intubation and mechanical ventilation (ETT-MV) during the first days of life to survive. Unfortunately this therapy is associated with adverse clinical outcomes and consequently, it is desirable to remove ETT-MV as quickly as possible. However, about 25% of extubated infants will fail and require re-intubation which is also associated with a 5-fold increase in mortality and a longer stay in the intensive care unit. Therefore, the ultimate goal is to determine the optimal time for extubation that will minimize the duration of MV and maximize the chances of success. This paper presents a new objective predictor to assist clinicians in making this decision. The predictor uses a modern machine learning method (Support Vector Machines) to determine the combination of measures of cardiorespiratory variability, computed automatically, that best predicts extubation readiness. Our results demonstrate that this predictor accurately classified infants who would fail extubation.

Previously, we presented automated methods for thoraco-abdominal asynchrony estimation and movement artifact detection in respiratory inductance plethysmography (RIP) signals. This paper combines and improves these methods to give a method for the automated, off-line detection of pause, movement artifact, and asynchrony. Simulation studies demonstrated that the new combined method is accurate and robust in the presence of noise. The new procedure was successfully applied to cardiorespiratory signals acquired postoperatively from infants in the recovery room. A comparison of the events detected with the automated method to those visually scored by an expert clinician demonstrated a higher agreement (κ = 0.52) than that amongst several human scorers (κ = 0.31) in a clinical study . The method provides the following advantages: first, it is fully automated; second, it is more efficient than visual scoring; third, the analysis is repeatable and standardized; fourth, it provides greater agreement with an expert scorer compared to the agreement between trained scorers; fifth, it is amenable to online detection; and lastly, it is applicable to uncalibrated RIP signals. Examples of applications include respiratory monitoring of postsurgical patients and sleep studies.

We recently presented a comprehensive automated off-line method for supervised respiratory event classification from uncalibrated respiratory inductive plethysmography signals. This method required training with a sample of clinical measurements classified by an expert. This human intervention is labor intensive and involves subjective judgments that may introduce bias to the automated classification. To address this we developed a novel method for unsupervised respiratory event classification, named AUREA (Automated Unsupervised Respiratory Event Analysis). This paper describes the algorithm underlying AUREA and demonstrates its successful application to respiratory signals acquired from infants in the postoperative recovery room. The advantages of AUREA are: first, it provides real-time classification of respiratory events; second, it requires no human intervention; and lastly, it has substantially better performance than the supervised method.