In Silico Design of Heterogeneous Microvascular Trees Using Generative Adversarial Networks and Constrained Constructive Optimization.

OBJECTIVE: Designing physiologically adequate microvascular trees is of crucial relevance for bioengineering functional tissues and organs. Yet, currently available methods are poorly suited to replicate the morphological and topological heterogeneity of real microvascular trees because the parameters used to control tree generation are too simplistic to mimic results of the complex angiogenetic and structural adaptation processes in vivo. METHODS: We propose a method to overcome this limitation by integrating a conditional deep convolutional generative adversarial network (cDCGAN) with a local fractal dimension-oriented constrained constructive optimization (LFDO-CCO) strategy. The cDCGAN learns the patterns of real microvascular bifurcations allowing for their artificial replication. The LFDO-CCO strategy connects the generated bifurcations hierarchically to form microvascular trees with a vessel density corresponding to that observed in healthy tissues. RESULTS: The generated artificial microvascular trees are consistent with real microvascular trees regarding characteristics such as fractal dimension, vascular density, and coefficient of variation of diameter, length, and tortuosity. CONCLUSIONS: These results support the adoption of the proposed strategy for the generation of artificial microvascular trees in tissue engineering as well as for computational modeling and simulations of microcirculatory physiology.

Reinforcement Learning Model for Managing Noninvasive Ventilation Switching Policy.

Noninvasive ventilation (NIV) has been recognized as a first-line treatment for respiratory failure in patients with chronic obstructive pulmonary disease (COPD) and hypercapnia respiratory failure, which can reduce mortality and burden of intubation. However, during the long-term NIV process, failure to respond to NIV may cause overtreatment or delayed intubation, which is associated with increased mortality or costs. Optimal strategies for switching regime in the course of NIV treatment remain to be explored.For the goal of reducing 28-day mortality of the patients undergoing NIV, Double Dueling Deep Q Network (D3QN) of offline-reinforcement learning algorithm was adopted to develop an optimal regime model for making treatment decisions of discontinuing ventilation, continuing NIV, or intubation. The model was trained and tested using the data from Multi-Parameter Intelligent Monitoring in Intensive Care III (MIMIC-III) and evaluated by the practical strategies. Furthermore, the applicability of the model in majority disease subgroups (Catalogued by International Classification of Diseases, ICD) was investigated. Compared with physician's strategies, the proposed model achieved a higher expected return score (4.25 vs. 2.68) and its recommended treatments reduced the expected mortality from 27.82% to 25.44% in all NIV cases. In particular, for these patients finally received intubation in practice, if the model also supported the regime, it would warn of switching to intubation 13.36 hours earlier than clinicians (8.64 vs. 22 hours after the NIV treatment), granting a 21.7% reduction in estimated mortality. In addition, the model was applicable across various disease groups with distinguished achievement in dealing with respiratory disorders. The proposed model is promising to dynamically provide personalized optimal NIV switching regime for patients undergoing NIV with the potential of improving treatment outcomes.

Comprehensive breathing variability indices enhance the prediction of extubation failure in patients on mechanical ventilation.

BACKGROUND AND OBJECTIVE: Despite the numerous studies on extubation readiness assessment for patients who are invasively ventilated in the intensive care unit, a 10-15% extubation failure rate persists. Although breathing variability has been proposed as a potential predictor of extubation failure, it is mainly assessed using simple statistical metrics applied to basic respiratory parameters. Therefore, the complex pattern of breathing variability conveyed by continuous ventilation waveforms may be underexplored. METHODS: Here, we aimed to develop novel breathing variability indices to predict extubation failure among invasively ventilated patients. First, breath-to-breath basic and comprehensive respiratory parameters were computed from continuous ventilation waveforms 1 h before extubation. Subsequently, the basic and advanced variability methods were applied to the respiratory parameter sequences to derive comprehensive breathing variability indices, and their role in predicting extubation failure was assessed. Finally, after reducing the feature dimensionality using the forward search method, the combined effect of the indices was evaluated by inputting them into the machine learning models, including logistic regression, random forest, support vector machine, and eXtreme Gradient Boosting (XGBoost). RESULTS: The coefficient of variation of the dynamic mechanical power per breath (CV-MPd[J/breath]) exhibited the highest area under the receiver operating characteristic curve (AUC) of 0.777 among the individual indices. Furthermore, the XGBoost model obtained the best AUC (0.902) by combining multiple selected variability indices. CONCLUSIONS: These results suggest that the proposed novel breathing variability indices can improve extubation failure prediction in invasively ventilated patients.

Pulsatility damping in the microcirculation: Basic pattern and modulating factors.

Blood flow pulsatility is an important determinant of macro- and microvascular physiology. Pulsatility is damped largely in the microcirculation, but the characteristics of this damping and the factors that regulate it have not been fully elucidated yet. Applying computational approaches to real microvascular network geometry, we examined the pattern of pulsatility damping and the role of potential damping factors, including pulse frequency, vascular viscous resistance, vascular compliance, viscoelastic behavior of the vessel wall, and wave propagation and reflection. To this end, three full rat mesenteric vascular networks were reconstructed from intravital microscopic recordings, a one-dimensional (1D) model was used to reproduce pulsatile properties within the network, and potential damping factors were examined by sensitivity analysis. Results demonstrate that blood flow pulsatility is predominantly damped at the arteriolar side and remains at a low level at the venular side. Damping was sensitive to pulse frequency, vascular viscous resistance and vascular compliance, whereas viscoelasticity of the vessel wall or wave propagation and reflection contributed little to pulsatility damping. The present results contribute to our understanding of mechanical forces and their regulation in the microcirculation.

Assessment of respiratory system compliance under pressure control ventilation without an inspiratory pause maneuver.

Objective. The measurement of the static compliance of the respiratory system (Cstat) during mechanical ventilation requires zero end-inspiratory flow. An inspiratory pause maneuver is needed if the zero end-inspiratory flow condition cannot be satisfied under normal ventilation.Approach. We propose a method to measure the quasi-static respiratory compliance (Cqstat) under pressure control ventilation mode without the inspiratory pause maneuver. First, a screening strategy was applied to filter out breaths affected strongly by spontaneous breathing efforts or artifacts. Then, we performed a virtual extrapolation of the flow-time waveform when the end-inspiratory flow was not zero, to allow for the calculation ofCqstatfor each kept cycle. Finally, the outputCqstatwas obtained as the average of the smallest 40Cqstatmeasurements. The proposed method was validated against the gold standardCstatmeasured from real clinical settings and compared with two reported algorithms. The gold standardCstatwas obtained by applying an end-inspiratory pause maneuver in the volume-control ventilation mode.Main results. Sixty-nine measurements from 36 patients were analyzed. The Bland-Altman analysis showed that the bias of agreement forCqstatversus the gold standard measurement was -0.267 ml/cmH2O (95% limits of agreement was -4.279 to 4.844 ml/cmH2O). The linear regression analysis indicated a strong correlation (R2 = 0.90) between theCqstatand gold standard.Significance. The results showed that theCqstatcan be accurately estimated from continuous ventilator waveforms, including spontaneous breathing without an inspiratory pause maneuver. This method promises to provide continuous measurements compliant with mechanical ventilation.

Identifying Patient-Ventilator Asynchrony on a Small Dataset Using Image-Based Transfer Learning.

Mechanical ventilation is an essential life-support treatment for patients who cannot breathe independently. Patient-ventilator asynchrony (PVA) occurs when ventilatory support does not match the needs of the patient and is associated with a series of adverse clinical outcomes. Deep learning methods have shown a strong discriminative ability for PVA detection, but they require a large number of annotated data for model training, which hampers their application to this task. We developed a transfer learning architecture based on pretrained convolutional neural networks (CNN) and used it for PVA recognition based on small datasets. The one-dimensional signal was converted to a two-dimensional image, and features were extracted by the CNN using pretrained weights for classification. A partial dropping cross-validation technique was developed to evaluate model performance on small datasets. When using large datasets, the performance of the proposed method was similar to that of non-transfer learning methods. However, when the amount of data was reduced to 1%, the accuracy of transfer learning was approximately 90%, whereas the accuracy of the non-transfer learning was less than 80%. The findings suggest that the proposed transfer learning method can obtain satisfactory accuracies for PVA detection when using small datasets. Such a method can promote the application of deep learning to detect more types of PVA under various ventilation modes.

An interpretable 1D convolutional neural network for detecting patient-ventilator asynchrony in mechanical ventilation.

BACKGROUND AND OBJECTIVE: Patient-ventilator asynchrony (PVA) is the result of a mismatch between the need of patients and the assistance provided by the ventilator during mechanical ventilation. Because the poor interaction between the patient and the ventilator is associated with inferior clinical outcomes, effort should be made to identify and correct their occurrence. Deep learning has shown promising ability in PVA detection; however, lack of network interpretability hampers its application in clinic. METHODS: We proposed an interpretable one-dimensional convolutional neural network (1DCNN) to detect four most manifestation types of PVA (double triggering, ineffective efforts during expiration, premature cycling and delayed cycling) under pressure control ventilation mode and pressure support ventilation mode. A global average pooling (GAP) layer was incorporated with the 1DCNN model to highlight the sections of the respiratory waveform the model focused on when making a classification. Dilation convolution and batch normalization were introduced to the 1DCNN model for compensating the reduction of performance caused by the GAP layer. RESULTS: The proposed interpretable 1DCNN exhibited comparable performance with the state-of-the-art deep learning model in PVA detection. The F1 scores for the detection of four types of PVA under pressure control ventilation and pressure support ventilation modes were greater than 0.96. The critical sections of the waveform used to detect PVA were highlighted, and found to be well consistent with the understanding of the respective type of PVA by experts. CONCLUSIONS: The findings suggest that the proposed 1DCNN can help detect PVA, and enhance the interpretability of the classification process to help clinicians better understand the results obtained from deep learning technology.

Complete mitochondrial genome and the phylogenetic position of a new species, Johnius taiwanensis (Perciformes: Sciaenidae) from Chinese waters.

In this study, the complete mitogenome of a new species, Johnius taiwanensis (Chao et al. 2019) was obtained. Its mitogenome is 18,451 bp in length, consisting of 37 genes with the typical gene order and direction of transcription in vertebrates. Gene rearrangement was found in J. taiwanensis. The overall nucleotide composition is: 24.2% A; 18.0% C; 21.1% G, and 36.7% T. Sizes of the 22 tRNA genes range from 66 to 75 bp. Two start codons (ATG and GTG) and three stop codons (TAG, AGA and TAA/TA/T) were detected in 13 protein-coding genes. In the Bayesian tree based on the complete mitogenomes of 26 species (including J. taiwanensis) from the family Sciaenidae, all nodes were strongly supported. The result shows that J. taiwanensis was placed as sister to the Trewavas croaker J. trewavasae of the same genus. The mechanism of gene rearrangement in the genus Johnius merits further investigation.

Characteristics on the oxidation stability of infant formula powder with different ingredients during storage.

Infant formula powder is prone to oxidation reaction during storage, which leads to the decrease of milk powder quality. The whole milk powder (WMP) was formulated, and the characteristics of infant formula powder were tracked during storage. The addition of metal ions, polyunsaturated fatty acids, and vitamins could reduce the peroxide value and increase the thiobarbituric acid value in the infant formula powder during the early stage of storage. When the samples were stored for 6 months, the free fat content of the base milk powder and the sample added with metal ions had high level (3.3%-3.6%). With adding vitamins, the content of free fat in the samples decreased first and then increased. The color value L of all the samples decreased during storage. Compared with WMP, the color value B of all the infant formula powder with different ingredients decreased. Levels of 2-heptanone and 2-nonaone indicated that the formation of the main methyl ketones in the infant formula powder with different ingredients decreased. The content of hexanal in the sample added metal ions was the highest. The type and intensity of free radicals changed with the formula components. The range of g value was 2.0043-2.0060 after 6 months of storage and 2.0017-2.1338 after 12 months of storage, respectively. The index of peroxide value and color value B were significantly related to the existence of free radicals in the infant formula powder with different ingredients.

Risk Factors for Patient-Ventilator Asynchrony and Its Impact on Clinical Outcomes: Analytics Based on Deep Learning Algorithm.

Background and objectives: Patient-ventilator asynchronies (PVAs) are common in mechanically ventilated patients. However, the epidemiology of PVAs and its impact on clinical outcome remains controversial. The current study aims to evaluate the epidemiology and risk factors of PVAs and their impact on clinical outcomes using big data analytics. Methods: The study was conducted in a tertiary care hospital; all patients with mechanical ventilation from June to December 2019 were included for analysis. Negative binomial regression and distributed lag non-linear models (DLNM) were used to explore risk factors for PVAs. PVAs were included as a time-varying covariate into Cox regression models to investigate its influence on the hazard of mortality and ventilator-associated events (VAEs). Results: A total of 146 patients involving 50,124 h and 51,451,138 respiratory cycles were analyzed. The overall mortality rate was 15.6%. Double triggering was less likely to occur during day hours (RR: 0.88; 95% CI: 0.85-0.90; p < 0.001) and occurred most frequently in pressure control ventilation (PCV) mode (median: 3; IQR: 1-9 per hour). Ineffective effort was more likely to occur during day time (RR: 1.09; 95% CI: 1.05-1.13; p < 0.001), and occurred most frequently in PSV mode (median: 8; IQR: 2-29 per hour). The effect of sedatives and analgesics showed temporal patterns in DLNM. PVAs were not associated mortality and VAE in Cox regression models with time-varying covariates. Conclusions: Our study showed that counts of PVAs were significantly influenced by time of the day, ventilation mode, ventilation settings (e.g., tidal volume and plateau pressure), and sedatives and analgesics. However, PVAs were not associated with the hazard of VAE or mortality after adjusting for protective ventilation strategies such as tidal volume, plateau pressure, and positive end expiratory pressure (PEEP).

Preprocessing Unevenly Sampled RR Interval Signals to Enhance Estimation of Heart Rate Deceleration and Acceleration Capacities in Discriminating Chronic Heart Failure Patients from Healthy Controls.

Objective. The deceleration capacity (DC) and acceleration capacity (AC) of heart rate, which are recently proposed variants to the heart rate variability, are calculated from unevenly sampled RR interval signals using phase-rectified signal averaging. Although uneven sampling of these signals compromises heart rate variability analyses, its effect on DC and AC analyses remains to be addressed. Approach. We assess preprocessing (i.e., interpolation and resampling) of RR interval signals on the diagnostic effect of DC and AC from simulation and clinical data. The simulation analysis synthesizes unevenly sampled RR interval signals with known frequency components to evaluate the preprocessing performance for frequency extraction. The clinical analysis compares the conventional DC and AC calculation with the calculation using preprocessed RR interval signals on 24-hour data acquired from normal subjects and chronic heart failure patients. Main Results. The assessment of frequency components in the RR intervals using wavelet analysis becomes more robust with preprocessing. Moreover, preprocessing improves the diagnostic ability based on DC and AC for chronic heart failure patients, with area under the receiver operating characteristic curve increasing from 0.920 to 0.942 for DC and from 0.818 to 0.923 for AC. Significance. Both the simulation and clinical analyses demonstrate that interpolation and resampling of unevenly sampled RR interval signals improve the performance of DC and AC, enabling the discrimination of CHF patients from healthy controls.

Detection of patient-ventilator asynchrony from mechanical ventilation waveforms using a two-layer long short-term memory neural network.

BACKGROUND AND OBJECTIVE: Mismatch between invasive mechanical ventilation and the requirements of patients results in patient-ventilator asynchrony (PVA), which is associated with a series of adverse clinical outcomes. Although the efficiency of the available approaches for automatically detecting various types of PVA from the ventilator waveforms is unsatisfactory, the feasibility of powerful deep learning approaches in addressing this problem has not been investigated. METHODS: We propose a 2-layer long short-term memory (LSTM) network to detect two most frequently encountered types of PVA, namely, double triggering (DT) and ineffective inspiratory effort during expiration (IEE), on two datasets. The performance of the networks is evaluated first using cross-validation on the combined dataset, and then using a cross testing scheme, in which the LSTM networks are established on one dataset and tested on the other. RESULTS: Compared with the reported rule-based algorithms and the machine learning models, the proposed 2-layer LSTM network exhibits the best overall performance, with the F1 scores of 0.983 and 0.979 for DT and IEE detection, respectively, on the combined dataset. Furthermore, it outperforms the other approaches in cross testing. CONCLUSIONS: The findings suggest that LSTM is an excellent technique for accurate recognition of PVA in clinics. Such a technique can help detect and correct PVA for a better patient ventilator interaction.

Arenibacter aquaticus sp. nov., a marine bacterium isolated from surface sea water in the South China Sea.

A Gram-stain-negative and facultatively anaerobic bacterial strain, designated GUOT, was isolated from surface water collected from the South China Sea. Cells were non-flagellate, yellow, non-spore-forming and rod-shaped. The 16S rRNA gene sequence comparisons with species in the genus Arenibacter showed that strain GUOT shares the highest similarity of 97.5 % with Arenibacter echinorum and Arenibacter palladensis. Average nucleotide identity and digital DNA-DNA hybridization values between strain GUOT and its related type strains were 77.1-78.4% and 20.8-26.2 % respectively. Growth of strain GUOT occurred at 15-50°C (optimum, 20-25°C), pH 5-7.5 (pH 6) and in media containing 0-7 % NaCl (optimum, 0-1 %). Cells contained methanol-soluble yellow-coloured pigments but flexirubin-type pigments were absent. The major fatty acids (>5 %) were iso-C17 : 0 3-OH, iso-C15 : 0, anteiso-C15 : 0, C16 : 0, summed feature 3, iso-C15 : 1 G and iso-C15 : 0 3-OH. The dominant polar lipids comprised phosphatidylethanolamine and some unidentified polar lipids. The main respiratory quinone was menaquinone-6. The DNA G+C content of strain GUOT was 40.1 %. Based on the presented data, we consider strain GUOT to represent a novel species of the genus Arenibacter, for which the name Arenibacter aquaticus sp. nov. is proposed. The type strain is GUOT (=KCTC 62629T=MCCC 1K03559T).

Reproductive Biology of Johnius taiwanensis (Perciformes: Sciaenidae) in Fujian Waters, Southern China.

Johnius taiwanensis is a newly described species from the Family Sciaenidae (Perciformes). The species is commonly found in shallow coastal waters along both sides of the Taiwan Strait, on the west sides of Zhejiang, Fujian, Guangdong and Hong Kong and east side of Taiwan, and has been misidentified for decades. We studied the reproductive biology of J. taiwanensis from Fujian coastal waters, southern China, using gonadosomatic index (GSI) and gonad histology analyses. Monthly sampling from July 2016 to October 2017 was conducted and a total of 638 specimens were collected, ranging from 7.3 to 19.0 cm standard length (SL). Gonad histology suggested that the spawning activity of J. taiwanensis females and males lasted from April to October, and the peak spawning months for females was July to September. Mature females and males were 12.5 and 11.8 cm SL, respectively, while the estimated sizes at 50% maturity were 12.0 cm and 10.9 cm SL, respectively. Vitellogenic stage oocytes (O3) and post-ovulatory follicles (POF) or hydrated oocytes (HO) were observed, and POF and O3 in ovaries indicated that J. taiwanensis spawns multiple times each spawning season. HO or both HO and POF were observed in ovaries collected from one same location in May 2017 and August 2016.

Pulse wave velocity in the microcirculation reflects both vascular compliance and resistance: Insights from computational approaches.

OBJECTIVE: PWV is the speed of pulse wave propagation through the circulatory system. mPWV emerges as a novel indicator of hypertension, yet it remains unclear how different vascular properties affect mPWV. We aim to identify the biomechanical determinants of mPWV. METHODS: A 1D model was used to simulate PWV in a rat mesenteric microvascular network and, for comparison, in a human macrovascular arterial network. Sensitivity analysis was performed to assess the relationship between PWV and vascular compliance and resistance. RESULTS: The 1D model enabled adequate simulation of PWV in both micro- and macrovascular networks. Simulated arterial PWV changed as a function of vascular compliance but not resistance, in that arterial PWV varied at a rate of 0.30 m/s and -6.18 × 10-3  m/s per 10% increase in vascular compliance and resistance, respectively. In contrast, mPWV depended on both vascular compliance and resistance, as it varied at a rate of 2.79 and -2.64 cm/s per 10% increase in the respective parameters. CONCLUSIONS: The present study identifies vascular compliance and resistance in microvascular networks as critical determinants of mPWV. We anticipate that mPWV can be utilized as an effective indicator for the assessment of microvascular biomechanical properties.

[Optimization of the parameters of microcirculatory structural adaptation model based on improved quantum-behaved particle swarm optimization algorithm].

The vessels in the microcirculation keep adjusting their structure to meet the functional requirements of the different tissues. A previously developed theoretical model can reproduce the process of vascular structural adaptation to help the study of the microcirculatory physiology. However, until now, such model lacks the appropriate methods for its parameter settings with subsequent limitation of further applications. This study proposed an improved quantum-behaved particle swarm optimization (QPSO) algorithm for setting the parameter values in this model. The optimization was performed on a real mesenteric microvascular network of rat. The results showed that the improved QPSO was superior to the standard particle swarm optimization, the standard QPSO and the previously reported Downhill algorithm. We conclude that the improved QPSO leads to a better agreement between mathematical simulation and animal experiment, rendering the model more reliable in future physiological studies.

The degree of heart rate asymmetry is crucial for the validity of the deceleration and acceleration capacity indices of heart rate: A model-based study.

The deceleration capacity (DC) and acceleration capacity (AC) of heart rate are a pair of indices used for evaluating the autonomic nervous system (ANS). We assessed the role of heart rate asymmetry (HRA) in defining the relative performance of DC and AC using a mathematical model, which is able to generate a realistic RR interval (RRI) time series with controlled ANS states. The simulation produced a set of RRI series with random sympathetic and vagal activities. The multi-scale DCs and ACs were computed from the RRI series, and the correlation of DC and AC with the ANS functions was analyzed to evaluate the performance of the indices. In the model, the HRA level was modified by changing the inspiration/expiration (I/E) ratio to examine the influence of HRA on the performances of DC and AC. The results show that on the conventional scales (T=1, s=2), an HRA level above 50% results in a stronger association of DC with the ANS, compared with AC. On higher scales (T=4, s=6), there was no HRA and DC showed a similar performance to AC for all I/E ratios. The data suggest that the HRA level determines which of DC or AC is the optimal index for expressing ANS functions. Future clinical applications of DC and AC should be accompanied by an HRA analysis to provide a better index for assessing ANS.

Do the deceleration/acceleration capacities of heart rate reflect cardiac sympathetic or vagal activity? A model study.

Despite increased application of the deceleration capacity (DC) and acceleration capacity (AC) of heart rate indices as indicators of autonomic nervous system (ANS) function, it remains controversial as to whether they reflect cardiac sympathetic or vagal activity. We addressed this problem using a cardiovascular system model that allows analysis of DC and AC under controllable levels of sympathetic and vagal activities. Multi-scale DCs and ACs with various timescales T and wavelet scales s were computed from the simulated RR interval series under randomly fluctuating levels of ANS activity, and the correlations of the indices to ANS functions were assessed. Results showed that under the conventional scales (T = 1, s = 2), both DC and AC were solely dependent on vagal activity. With higher scales (T = 3, s = 5), both DC and AC were positively correlated to sympathetic activity and negatively correlated to vagal activity. These data suggest that DC and AC provide information on the same aspects of ANS activity and that their physiological significance is highly influenced by the timescales and wavelet scales used in the computation.

Simulation of microcirculatory hemodynamics: estimation of boundary condition using particle swarm optimization.

Estimation of the boundary condition is a critical problem in simulating hemodynamics in microvascular networks. This paper proposed a boundary estimation strategy based on a particle swarm optimization (PSO) algorithm, which aims to minimize the number of vessels with inverted flow direction in comparison to the experimental observation. The algorithm took boundary values as the particle swarm and updated the position of the particles iteratively to approach the optimization target. The method was tested in a real rat mesenteric network. With random initial boundary values, the method achieved a minimized 9 segments with an inverted flow direction in the network with 546 vessels. Compared with reported literature, the current work has the advantage of a better fit with experimental observations and is more suitable for the boundary estimation problem in pulsatile hemodynamic models due to the experiment-based optimization target selection.

[A new method of measuring the pulse based on facial video].

This paper proposes a new method of non-contact pulse measurement by analyzing a clip of human facial video. The method is based on photo plethysmography (PPG) and independent component analysis (ICA) model. A clip of color facial video shot under normal lighting condition is firstly discomposed into RGB channel sequences. Secondly, by applying ICA to the 3 channel sequences, 3 new independent signals are obtained, among which one signal is close to human pulse wave. Thus the pulse can be measured. In this paper, the principles of PPG and ICA are briefly described and the measurement framework is proposed. The experimental results showed that this novel approach was reasonable and feasible.