Prediction of extubation readiness in extreme preterm infants based on measures of cardiorespiratory variability

Gupta

Numer Methods Biomed Eng

2018

Infants born prematurely are particularly susceptible to respiratory illness due to underdeveloped lungs, which can often result in fatality. Preterm infants in acute stages of respiratory illness typically require mechanical ventilation assistance, and the efficacy of the type of mechanical ventilation and its delivery has been the subject of a number clinical studies. With recent advances in machine learning approaches, particularly deep learning, it may be possible to estimate future responses to mechanical ventilation in real time, based on ventilation monitoring up to the point of analysis. In this work, recurrent neural networks are proposed for predicting future ventilation parameters due to the highly nonlinear behavior of the ventilation measures of interest and the ability of recurrent neural networks to model complex nonlinear functions. The resulting application of this particular class of neural networks shows promise in its ability to predict future responses for different ventilation modes. Towards improving care and treatment of preterm newborns, further development of this prediction process for ventilation could potentially aid in important clinical decisions or studies to improve preterm infant health.

Section: Introductionmentioning

confidence: 99%

Predicting responses to mechanical ventilation for preterm infants with acute respiratory illness using artificial neural networks

Gupta

Numer Methods Biomed Eng

2018

2017 IEEE Symposium Series on Computational Intelligence (SSCI)

“…This means that at the time infants were deemed ready for extubation, 8 out of 10 failures would be prevented, but at the expense of unnecessarily prolonging the course of mechanical ventilation in 5 out of 10 successes. Previous work [12] which applied SVM (on a much smaller cohort) using cardiorespiratory variables gave comparable sensitivity but a higher specificity of 74%. It will be necessary to repeat those experiments with this larger dataset.…”

Section: Discussionmentioning

confidence: 91%

“…high specificity), but not as sensitive in detecting failures. In [12], using patients of the same cohort, a Support Vector Machine (SVM) was developed to predict extubation readiness directly from the combination of cardiorespiratory variability measures. This system achieved improved ability to detect extubation failures (sensitivity of 83%) while maintaining a fairly high specificity of 74%.…”

Section: Related Workmentioning

confidence: 99%

Predicting extubation readiness in extreme preterm infants based on patterns of breathing

Onu

Kanbar

Shalish

et al. 2017

Extremely preterm infants commonly require intubation and invasive mechanical ventilation after birth. While the duration of mechanical ventilation should be minimized in order to avoid complications, extubation failure is associated with increases in morbidities and mortality. As part of a prospective observational study aimed at developing an accurate predictor of extubation readiness, Markov and semi-Markov chain models were applied to gain insight into the respiratory patterns of these infants, with more robust timeseries modeling using semi-Markov models. This model revealed interesting similarities and differences between newborns who succeeded extubation and those who failed. The parameters of the model were further applied to predict extubation readiness via generative (joint likelihood) and discriminative (support vector machine) approaches. Results showed that up to 84% of infants who failed extubation could have been accurately identified prior to extubation.

2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

“…Abbreviated cardiorespiratory studies (ACS) are commonly performed to assess/study respiratory conditions such as apnea, when full polysomnography (PSG) is not indicated or available (e.g., in the home [1], the recovery room [2], or the intensive care unit [3]). These studies acquire a subset of the PSG signals, typically including blood oxygen saturation (SaO 2 ) and photoplethysmography (PPG), measured with an oximeter, as well as the ribcage and abdomen respiratory movements from respiratory inductive plethysmography (RIP).…”

Section: Introductionmentioning

confidence: 99%

A new movement artifact detector for photoplethysmographic signals

Robles-Rubio

Brown

Kearney

2013

Self Cite

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.