Neil R. Euliano

Monitoring uterine activity during labor: a comparison of 3 methods

OBJECTIVE Tocodynamometry (Toco; strain gauge technology) provides contraction frequency and approximate duration of labor contractions but suffers frequent signal dropout, necessitating repositioning by a nurse, and may fail in obese patients. The alternative invasive intrauterine pressure catheter (IUPC) is more reliable and adds contraction pressure information but requires ruptured membranes and introduces small risks of infection and abruption. Electrohysterography (EHG) reports the electrical activity of the uterus through electrodes placed on the maternal abdomen. This study compared all 3 methods of contraction detection simultaneously in laboring women. STUDY DESIGN Upon consent, laboring women were monitored simultaneously with Toco, EHG, and IUPC. Contraction curves were generated in real-time for the EHG, and all 3 curves were stored electronically. A contraction detection algorithm was used to compare frequency and timing between methods. Seventy-three subjects were enrolled in the study; 14 were excluded due to hardware failure of 1 or more of the devices (n = 12) or inadequate data collection duration (n = 2). RESULTS In comparison with the gold-standard IUPC, EHG performed significantly better than Toco with regard to the Contractions Consistency Index (CCI). The mean CCI for EHG was 0.88 ± 0.17 compared with 0.69 ± 0.27 for Toco (P < .0001). In contrast to Toco, EHG was not significantly affected by obesity. CONCLUSION Toco does not correlate well with the gold-standard IUPC and fails more frequently in obese patients. EHG provides a reliable noninvasive alternative, regardless of body habitus.

Spatiotemporal electrohysterography patterns in normal and arrested labor

OBJECTIVE The purpose of this study was to investigate the spatiotemporal patterns of uterine electrical activity in normal and arrested labors. STUDY DESIGN From a database of electrohysterograms, 12 subjects who underwent cesarean delivery for active-phase arrest were each matched with 2 vaginally delivered controls. Using 30-minute segments of the electrohysterogram during the arrest, or the same dilation in controls, the center of uterine electrical activity was derived. The vertical motion of this center of uterine activity was determined for each contraction and the frequencies of movement patterns analyzed. RESULTS Predominantly upward movement of the center of uterine activity (longer and/or stronger contraction at the fundus) was more common with normal dilation (P = .003). Receiver operating characteristic curve analysis gave an area under the curve of 0.91 for predicting outcome (vaginal vs cesarean delivery). CONCLUSION There is a significant correlation between upward movement of the center of uterine activity (fundal dominance) and current labor progress.

Spatio-temporal self-organizing feature maps

Thus far, the success of capturing and classifying temporal information with neural networks has been limited. Our methodology adds a spatio-temporal coupling to the self-organizing feature map (SOFM) which creates temporally and spatially localized neighborhoods in the map. The spatio-temporal coupling is based on traveling waves of activity which start at each winning node and are naturally attenuated over time. When these traveling waves reinforce each other, temporal activity wavefronts are created which are then used to enhance a nodes possibility of winning the next competition. The spatiotemporal coupling is easily implemented with only local connectivity and calculations. Once trained, the spatio-temporal SOFM can be used for detection or for partial pattern recall. The methodology gracefully handles time-warping and multiple patterns with overlapping input vectors.

Independent components analysis for fetal electrocardiogram extraction: a case for the data efficient Mermaid algorithm

Fetal heart rate (FHR) monitoring is currently the primary methodology for antenatal determination of fetal well-being. Currently, the FHR can be detected with ultrasonography, but the additional information from fetal electrocardiogram (FECG) is only available via an invasive scalp electrode. A cost effective noninvasive monitoring through standard ECG electrodes could be used on nearly every patient in lieu of the ultrasound monitors. In this method, a number of electrodes are positioned on the abdomen of the mother to collect, simultaneously, various combinations of the signals including the heartbeats of the mother and the fetus. For accurate fetal heart-rate estimation, a clean FECG must be extracted from the collected mixtures. It is well known that this can be achieved using blind source separation (BSS) techniques. In this paper, the performance of the Mermaid algorithm, which is based on minimizing Renyis mutual information, is evaluated on this problem of great practical importance. The effectiveness and data efficiency of Mermaid and its superiority over alternative information theoretic BSS algorithms are illustrated using artificially mixed ECG signals as well as fetal heart rate estimates in real ECG mixtures.

A Spatio-Temporal Memory Based on SOMs with Activity Diffusion

Publisher Summary This chapter explores that sensory processing can be grouped into two domains—static and dynamic problems. Static problems consist of information that is independent of time. For instance, in static image recognition, the image does not change over time. On the other hand, time is fundamental to the dynamic problem. The output of a dynamical system, for example, depends not only on the present input but also on the current state of the system that encapsulates the entire past of the inputs. Temporal processing is the analysis, modeling, prediction, and/or classification of systems that vary with time. Patterns that evolve over time have traditionally provided the most challenging problems for scientists and engineers. This chapter discusses the use of the biologically inspired concept of activity diffusion to create a spatio-temporal memory in the self-organizing map (SOM) and neural gas algorithms. The activity diffusion creates a system that is sensitive to the temporal patterns it has been trained with, and thus can “anticipate” future inputs. This technique uses temporal information to help remove variability inherent in the signal.

Principles and networks for self-organization in space-time

In this paper, we develop a spatio-temporal memory that blends properties from long and short-term memory and is motivated by reaction diffusion mechanisms. The winning processing element of a self-organizing network creates traveling waves on the output space that gradually attenuate over time and space to diffuse temporal information and create localized spatio-temporal neighborhoods for clustering. The novelty of the model is in the creation of time varying Voronoi tessellations anticipating the learned input signal dynamics even when the cluster centers are fixed. We test the method in a robot navigation task and in vector quantization of speech. This method performs better than conventional static vector quantizers based on the same data set and similar training conditions.

Expiratory time constant for determinations of plateau pressure, respiratory system compliance, and total resistance.

IntroductionWe hypothesized the expiratory time constant (ƬE) may be used to provide real time determinations of inspiratory plateau pressure (Pplt), respiratory system compliance (Crs), and total resistance (respiratory system resistance plus series resistance of endotracheal tube) (Rtot) of patients with respiratory failure using various modes of ventilatory support.MethodsAdults (n = 92) with acute respiratory failure were categorized into four groups depending on the mode of ventilatory support ordered by attending physicians, i.e., volume controlled-continuous mandatory ventilation (VC-CMV), volume controlled-synchronized intermittent mandatory ventilation (VC-SIMV), volume control plus (VC+), and pressure support ventilation (PSV). Positive end expiratory pressure as ordered was combined with all aforementioned modes. Pplt, determined by the traditional end inspiratory pause (EIP) method, was combined in equations to determine Crs and Rtot. Following that, the ƬE method was employed, ƬE was estimated from point-by-point measurements of exhaled tidal volume and flow rate, it was then combined in equations to determine Pplt, Crs, and Rtot. Both methods were compared using regression analysis.ResultsƬE, ranging from mean values of 0.54 sec to 0.66 sec, was not significantly different among ventilatory modes. The ƬE method was an excellent predictor of Pplt, Crs, and Rtot for various ventilatory modes; r2 values for the relationships of ƬE and EIP methods ranged from 0.94 to 0.99 for Pplt, 0.90 to 0.99 for Crs, and 0.88 to 0.94 for Rtot (P <0.001). Bias and precision values were negligible.ConclusionsWe found the ƬE method was just as good as the EIP method for determining Pplt, Crs, and Rtot for various modes of ventilatory support for patients with acute respiratory failure. It is unclear if the ƬE method can be generalized to patients with chronic obstructive lung disease. ƬE is determined during passive deflation of the lungs without the need for changing the ventilatory mode and disrupting a patients breathing. The ƬE method obviates the need to apply an EIP, allows for continuous and automatic surveillance of inspiratory Pplt so it can be maintained ≤ 30 cm H2O for lung protection and patient safety, and permits real time assessments of pulmonary mechanics.

Power of breathing determined noninvasively with use of an artificial neural network in patients with respiratory failure.

Objective:To determine work of breathing per minute or power of breathing noninvasively (POBN) by using an artificial neural network (ANN) without the need for an esophageal catheter in patients with respiratory failure. Design:Prospective study comparing the relationship between POBN and invasively measured power of breathing (POBI). Setting:Intensive care unit of a university hospital. Patients:Forty-five intubated adults (age, 51 ± 11 yrs; weight, 71 ± 18 kg; 28 males and 17 females) receiving pressure support ventilation (PSV). Interventions:Data from an esophageal catheter and airway pressure/flow sensor were used to measure POBI. A pretrained ANN provided real time calculation of POBN. POBI and POBN were measured at various levels of PSV, ranging from 5 to 25 cm H2O. Measurements and Main Results:POBN was highly correlated with POBI (r = 0.91; p < .002), and because POBN explained or predicted 83% of the variance in POBI, it was considered a very good predictor (r2 = 0.83; p < .002). Bias was negligible (0.00) and precision was clinically acceptable (2.2 J/min). Conclusions:POB can be calculated noninvasively with reasonable clinical accuracy for patients receiving ventilatory support by using an ANN. This method obviates the need for inserting an esophageal catheter and thus greatly simplifies measurement of POB. POBN may be a clinically useful tool for consideration when setting PSV to unload the respiratory muscles. Before considering its use in clinical practice, POBN would need to be incorporated within the context of load tolerance and shown to improve outcomes.

Dynamic subgrouping in RTRL provides a faster O(N/sup 2/) algorithm

Static grouping of processing elements (PEs) has been proposed to reduce the computational complexity of real time recurrent learning (RTRL) from O(n/sup 4/) to O(n/sup 2/), but performance suffers. This paper proposes a dynamic subgrouping of PEs estimated from a local approximation of the /spl pi/ matrix based on temporal Hebbian of sensitivities during training. The method is O(n/sup 2/) and leads to better performance.

Monitoring Uterine Activity during Labor: Clinician Interpretation of Electrohysterography versus Intrauterine Pressure Catheter and Tocodynamometry

Objective The aim of this article was to compare clinical interpretation of uterine activity tracings acquired by tocodynamometry and electrohysterography with the gold standard, intrauterine pressure. Study Design Using data from a previous study, subjects who had simultaneous monitoring with all three uterine activity devices were included in this study. These were parturients who required intrauterine pressure catheter (IUPC) placement for obstetric indication. A Web-based application displayed scrolling 30-minute segments of uterine activity. Two blinded obstetricians and two blinded obstetric nurses independently reviewed the segments, marking uninterpretable segments and the peak of each contraction. Interpretability was compared using positive percent agreement. False positives are contractions marked in the noninvasive strip that have no corresponding contraction in the IUPC strip. False negatives are the reverse. Results A total of 135 segments, acquired during either Stage 1 (active labor) or Stage 2 (pushing), from 105 women, were included in this analysis. For all four observers, both interpretability and sensitivity of electrohysterography exceeded that of tocodynamometry (p < 0.0001). This remained true for the obese population (96 segments). Conclusion Compared with the IUPC, electrohysterography is more sensitive and provides tracings that are more often interpretable than tocodynamometry for intrapartum monitoring; electrohysterography is also less affected by increasing maternal body mass index.