Prediction method for weaning outcomes based on machine learning and electrical impedance tomography / 医疗卫生装备

Objective To propose a method for predicting weaning outcomes based on machine learning and electrical impedance tomography(EIT).Methods Firstly,EIT image features were extracted from a total of 84 samples from 30 patients,and the important features screened with the extreme gradient boosting(XGBoost)algorithm were used as inputs to the model.Secondly,the prediction model was built with six machine learning methods,namely random forest(RF),support vector machines(SVM),XGBoost,gradient boosting decision tree(GBDT),logistic regression(LR)and decision tree(tree).Then the prediction model had its prediction performance evaluated by AUC,accuracy,sensitivity and specificity under imbalanced dataset,over-sampling balanced dataset and random under-sampling balanced dataset.Results In terms of AUC,accuracy and specificity,the model under the over-sampling balanced dataset and the random under-sampling balanced dataset behaved better than that under the imbalanced dataset(P＜0.05);in terms of sensitivity,the difference in model performance between the over-sampling balanced dataset and the imbalanced dataset was not statistically significant(P＞0.05),and the model performance under the random under-sampling balanced dataset decreased when compared with that under the imbalanced dataset(P＜0.05).There were no significant differences between the model performance under the over-sampling balanced dataset and that under the random under-sampling balanced dataset(P＞0.05).The model based on XGBoost behaved the best under the over-sampling balanced dataset,with an AUC of 0.769,an accuracy of 0.808,a sensitivity of 0.938 and a specificity of 0.600.Conclusion The method based on machine learning and EIT predicts weaning outcomes of patients with prolonged mechanical ventilation,and thus can be used for auxiliary decision support for clinicians to determine the appropriate timing of weaning.[Chinese Medical Equipment Journal,2023,44(10):1-6]

Positive End-expiratory Pressure Titration after Alveolar Recruitment Directed by Electrical Impedance Tomography / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>Electrical impedance tomography (EIT) is a real-time bedside monitoring tool, which can reflect dynamic regional lung ventilation. The aim of the present study was to monitor regional gas distribution in patients with acute respiratory distress syndrome (ARDS) during positive-end-expiratory pressure (PEEP) titration using EIT.</p><p><b>METHODS</b>Eighteen ARDS patients under mechanical ventilation in Department of Critical Care Medicine of Peking Union Medical College Hospital from January to April in 2014 were included in this prospective observational study. After recruitment maneuvers (RMs), decremental PEEP titration was performed from 20 cmH 2 O to 5 cmH 2 O in steps of 3 cmH 2 O every 5-10 min. Regional over-distension and recruitment were monitored with EIT.</p><p><b>RESULTS</b>After RMs, patient with arterial blood oxygen partial pressure (PaO 2) + carbon dioxide partial pressure (PaCO 2 ) >400 mmHg with 100% of fractional inspired oxygen concentration were defined as RM responders. Thirteen ARDS patients was diagnosed as responders whose PaO 2 + PaCO 2 were higher than nonresponders (419 ± 44 mmHg vs. 170 ± 73 mmHg, P < 0.0001). In responders, PEEP mainly increased recruited pixels in dependent regions and over-distended pixels in nondependent regions. PEEP alleviated global inhomogeneity of tidal volume and end-expiratory lung volume. PEEP levels without significant alveolar derecruitment and over-distension were identified individually.</p><p><b>CONCLUSIONS</b>After RMs, PEEP titration significantly affected regional gas distribution in lung, which could be monitored with EIT. EIT has the potential to optimize PEEP titration.</p>

Alveolar stability under different combinations of positive end-expiratory pressure and tidal volume: alveolar microscopy in isolated injured rat lungs / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>High positive end-expiratory pressure (PEEP) and low tidal volume (VT) ventilation is thought to be a protective ventilation strategy. It is hypothesized that the stabilization of collapsible alveoli during expiration contributes to lung protection. However, this hypothesis came from analysis of indirect indices like the analysis of the pressure-volume curve of the lung. The purpose of this study was to investigate isolated healthy and injured rat lungs by means of alveolar microscopy, in which combination of PEEP and VT is beneficial with respect to alveolar stability (I-E%).</p><p><b>METHODS</b>Alveolar stability was investigated in isolated, non-perfused mechanically ventilated rat lungs. Injured lungs were compared with normal lungs. For both groups three PEEP settings (5, 10, 20 cmH2O) were combined with three VT settings (6, 10, 15 ml/kg) resulting in nine PEEP-VT combinations per group. Analysis was performed by alveolar microscopy.</p><p><b>RESULTS</b>In normal lungs alveolar stability persisted in all PEEP-VT combinations (I-E% (3.2 +/- 11.0)%). There was no significant difference using different settings (P > 0.01). In contrast, alveoli in injured lungs were extremely instable at PEEP levels of 5 cmH2O (mean I-E% 100%) and 10 cmH2O (mean I-E% (30.7 +/- 16.8)%); only at a PEEP of 20 cmH2O were alveoli stabilized (mean I-E% of (0.2 +/- 9.3)%).</p><p><b>CONCLUSIONS</b>In isolated healthy lungs alveolar stability is almost unaffected by different settings of PEEP and VT. In isolated injured lungs only a high PEEP level of 20 cmH2O resulted in stabilized alveoli whereas lower PEEP levels are associated with alveolar instability.</p>