2010 Second International Conference on Intelligent Human-Machine Systems and Cybernetics, 2010
The indoor air pollution produced by the ventilating system of the central air-conditioning draws... more The indoor air pollution produced by the ventilating system of the central air-conditioning draws many people’s attention. In this paper, a new style of the central air-conditioning pipe aseptic sampling inspection robotic system is developed. The entire system comprises three parts: the caterpillar robot, computer and control box. Through real-time video transferred from cameras on robot, operators can remote control the robot in pipe to move flexibly, sample and high-pressure dust absorption effectively. The integral separation PID improved algorithm applied in the control system makes the robot in pipe travel straight accurately, ensuring the actual speed fluctuation deviation ratio range is in±5%..
2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), 2017
Deceleration capacity (DC) of heart rate, a novel indicator of autonomic nervous system, is calcu... more Deceleration capacity (DC) of heart rate, a novel indicator of autonomic nervous system, is calculated from the unevenly sampled RR tachogram. We aim to study whether the nature of uneven sampling of the RR tachogram affects the performance of DC. Interpolation and resampling procedures were applied to the RR tachogram to generate an evenly sampled RR interval (RRI) time series. In particular, the resampling frequency, the interpolating approach, and other relevant parameters were examined for optimizing the performance of the modified DC. DC calculated from the resampled RRI time series was compared to the conventional one computed from the raw RR tachogram. The conventional and modified DCs were applied in distinguishing the chronic heart failure (CHF) patients and healthy subjects. The results show that the modification improved the diagnostic accuracy of CHF from 85.11% to 90.43%, suggesting that the resampling procedure enhances DC in the assessment of CHF.
2016 Chinese Control and Decision Conference (CCDC), 2016
In this paper, the Direction basis function networks(DBFNs) are proposed to improve the problem o... more In this paper, the Direction basis function networks(DBFNs) are proposed to improve the problem of facial expression recognition. Due to the diversity and nonlinear characteristic of facial expression itself, We proved a new linear feature extraction approach of facial expression which achieved a good directional feature. Using Cohn-Kanade facial expression database, our method provides a better performance for facial expression recognition.
Background: It remains controversial as how to set positive end-expiratory pressure (PEEP) for ac... more Background: It remains controversial as how to set positive end-expiratory pressure (PEEP) for acute respiratory distress syndrome (ARDS) patients. This study aims to provide suggestions to the clinicians in selecting PEEP for ARDS patients receiving invasive mechanical ventilation based on artificial intelligence (AI).Methods: Invasively ventilated ARDS patients in MIMIC-IV and eICU databases were enrolled in the observational cohort study. An AI model trained by awarding survival for suggesting optimal PEEP was developed and tested on the MIMIC-IV database and externally validated on the eICU database. Three subgroups were defined in which the PEEP grades set by the AI model are lower, equal, and higher than that set by the clinicians (denoted as , , and , respectively). Intensive care unit (ICU) mortality and 28-day ventilation-free days are the primary and secondary outcomes.Results: 6839 (MIMIC-IV) and 2117 (eICU) ARDS admissions were included in the study. The ICU mortalities ...
Machine learning algorithms are susceptible to cyberattacks, posing security problems in computer... more Machine learning algorithms are susceptible to cyberattacks, posing security problems in computer vision, speech recognition, and recommendation systems. So far, researchers have made great strides in adopting adversarial training as a defensive strategy. Single-step adversarial training methods have been proposed as viable solutions for improving model generality and resilience. However, there has been little study to address this issue in the context of ownership-based recommendations, which may fail to capture stable results. In this work, we adapt the single-step adversarial training for ownership recommendation systems. Our main technical contributions are as follows: (1) We propose Adversarial Consumption and Production Relationship (ACPR), a model that combines factorization machine and single-step adversarial training for ownership recommendations. It enables us to take advantage of modeling consumption-production interactions with a factorization machine instead of the conventional matrix factorization method for ownership recommendations. (2) We enrich the ACPR technique with directional adversarial training and call our technique Adversarial Consumption and Production Relationship-Aware Directional Adversarial Model (ACPR-ADAM). The idea behind our ACPR-ADAM is that instead of the worst perturbation direction, the perturbation direction in the embedding space is restricted to other examples in the current embedding space, allowing us to incorporate the collaborative signal into the training process. Lastly, through extensive evaluations on Reddit and Pinterest, we demonstrate that our proposed method outperforms state-of-the-art methods. Compared with CPR and ACPR on Reddit and Pinterest datasets, our proposed ACPR-ADAM achieves 93%, 88%, and 72%, 69% enhancement in terms of AUC and HR, respectively. INDEX TERMS Adversarial attack, directional adversarial training, factorization machine, ownership recommendation.
Background and objectives: Patient–ventilator asynchronies (PVAs) are common in mechanically vent... more 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:...
2016 Computing in Cardiology Conference (CinC), 2016
Deceleration capacity (DC) of heart rate is a novel index for evaluating the activity of the auto... more Deceleration capacity (DC) of heart rate is a novel index for evaluating the activity of the autonomic nervous system (ANS). We examined whether controlling the inspiration/expiration (I/E) ratio benefits the DC analysis based on a model-generated RR interval (RRI) database. A cardiovascular system model was adopted to simulate RRI time series. The model allows analyzing the role of sympathetic and vagal activities in the ANS. The respiratory pattern can be controlled in the model. Three hundred RRI time series with random sympathetic and vagal activities were simulated. According to the ratio between the sympathetic and vagal activities (S/V ratio), these subjects were categorized into a case group (S/V>1) and a control group (S/V<1). DC was computed for each subject. The performance of DC in distinguishing the two groups was examined by the receiver operating characteristic (ROC) analysis. The respiratory period is set to 6 s. The I/E ratio was controlled as 1:2, 1:1 and 2:1...
Objective. The measurement of the static compliance of the respiratory system (C stat) during mec... more Objective. The measurement of the static compliance of the respiratory system (C stat) 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 (C qstat) 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 of C qstat for each kept cycle. Finally, the output C qstat was obtained as the average of the smallest 40 C qstat measurements. The proposed method was validated against the gold standard C stat measured from real clinical settings and compared with two reported algorithms. The gold standard C stat was 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 for C qstat versus 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 (R 2 = 0.90) between the C qstat and gold standard. Significance. The results showed that the C qstat can 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.
Machine learning classifiers are vulnerable to adversarial perturbation, and their presence raise... more Machine learning classifiers are vulnerable to adversarial perturbation, and their presence raises security concerns, especially in recommendation systems. While attacks and defense mechanisms in recommendation systems have received significant attention, Basic Iterative Method (BIM), which has been shown in Computer Vision to increase attack effectiveness by more than 60%, has received little attention in ownership recommendation. As a result, ownership recommender systems may be more sensitive to iterative perturbations, resulting in significant generalization errors. Adversarial Training, a regularization strategy that can withstand worst-case iterative perturbations, could be a viable option for improving model robustness and generalization. In this paper, we implement BIM for ownership recommendations. Through adversarial training, we propose the Adversarial Consumer and Producer Recommendation (ACPR) approach that integrates ownership features into a multi-objective pairwise ranking to capture the user’s preferences. The ACPR method learns a core embedding for each user and two transformation matrices that project the user’s core embedding into two role embeddings (i.e., a producer and consumer role) using an extension of matrix factorization. To minimize the impact of iterative perturbation, we train a consumer and producer recommender objective function using minimax adversarial training. Empirical studies on two Large-scale applications show that our method outperforms standard recommendation methods and recent methods that model ownership information.
Blood flow pulsatility is an important determinant of macro- and microvascular physiology. Pulsat... more 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.
Computer Methods and Programs in Biomedicine, 2021
BACKGROUND AND OBJECTIVE Patient-ventilator asynchrony (PVA) is the result of a mismatch between ... more 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.
Background and objective: Mismatch between invasive mechanical ventilation and the requirements o... more 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 2layer 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.
Automated detection of atrial fibrillation (AF) from electrocardiogram (ECG) recordings remains c... more Automated detection of atrial fibrillation (AF) from electrocardiogram (ECG) recordings remains challenging in real clinical settings. Deep neural networks (DNN) emerge as a promising tool for the task of AF detection. However, the success of DNN for AF detection is hampered by limited size and imbalanced number of samples in datasets. We propose a novel data augmentation strategy based on duplication, concatenation and resampling of ECG episodes to balance the number of samples among different categories as well as to increase the diversity of samples. The performance of the data augmentation method was examined on an AF database from Computing in Cardiology (CinC) challenge 2017. A 2-layer long short-term memory (LSTM) network was trained with the augmented dataset. Its ability of AF detection was evaluated using a 10-fold cross validation approach. And F1 score was adopted as the metrics. The AF detection results show that the proposed method was superior to two conventional data augmentation methods: window slicing and permutation. The network was also submitted to the evaluation system of the CinC challenge 2017. The F1 score obtained by the network using the proposed data augmentation method was close to the winner (0.82 vs. 0.83). In summary, the proposed data augmentation method provides an effective solution to enhance the dataset for improving the performance of DNN in ECG analysis. Such a method promotes the application of deep learning in the analysis of ECG, particularly when the dataset is small and imbalanced.
PWV is the speed of pulse wave propagation through the circulatory system. mPWV emerges as a nove... more 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. 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. 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 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. The...
2010 Second International Conference on Intelligent Human-Machine Systems and Cybernetics, 2010
The indoor air pollution produced by the ventilating system of the central air-conditioning draws... more The indoor air pollution produced by the ventilating system of the central air-conditioning draws many people’s attention. In this paper, a new style of the central air-conditioning pipe aseptic sampling inspection robotic system is developed. The entire system comprises three parts: the caterpillar robot, computer and control box. Through real-time video transferred from cameras on robot, operators can remote control the robot in pipe to move flexibly, sample and high-pressure dust absorption effectively. The integral separation PID improved algorithm applied in the control system makes the robot in pipe travel straight accurately, ensuring the actual speed fluctuation deviation ratio range is in±5%..
2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), 2017
Deceleration capacity (DC) of heart rate, a novel indicator of autonomic nervous system, is calcu... more Deceleration capacity (DC) of heart rate, a novel indicator of autonomic nervous system, is calculated from the unevenly sampled RR tachogram. We aim to study whether the nature of uneven sampling of the RR tachogram affects the performance of DC. Interpolation and resampling procedures were applied to the RR tachogram to generate an evenly sampled RR interval (RRI) time series. In particular, the resampling frequency, the interpolating approach, and other relevant parameters were examined for optimizing the performance of the modified DC. DC calculated from the resampled RRI time series was compared to the conventional one computed from the raw RR tachogram. The conventional and modified DCs were applied in distinguishing the chronic heart failure (CHF) patients and healthy subjects. The results show that the modification improved the diagnostic accuracy of CHF from 85.11% to 90.43%, suggesting that the resampling procedure enhances DC in the assessment of CHF.
2016 Chinese Control and Decision Conference (CCDC), 2016
In this paper, the Direction basis function networks(DBFNs) are proposed to improve the problem o... more In this paper, the Direction basis function networks(DBFNs) are proposed to improve the problem of facial expression recognition. Due to the diversity and nonlinear characteristic of facial expression itself, We proved a new linear feature extraction approach of facial expression which achieved a good directional feature. Using Cohn-Kanade facial expression database, our method provides a better performance for facial expression recognition.
Background: It remains controversial as how to set positive end-expiratory pressure (PEEP) for ac... more Background: It remains controversial as how to set positive end-expiratory pressure (PEEP) for acute respiratory distress syndrome (ARDS) patients. This study aims to provide suggestions to the clinicians in selecting PEEP for ARDS patients receiving invasive mechanical ventilation based on artificial intelligence (AI).Methods: Invasively ventilated ARDS patients in MIMIC-IV and eICU databases were enrolled in the observational cohort study. An AI model trained by awarding survival for suggesting optimal PEEP was developed and tested on the MIMIC-IV database and externally validated on the eICU database. Three subgroups were defined in which the PEEP grades set by the AI model are lower, equal, and higher than that set by the clinicians (denoted as , , and , respectively). Intensive care unit (ICU) mortality and 28-day ventilation-free days are the primary and secondary outcomes.Results: 6839 (MIMIC-IV) and 2117 (eICU) ARDS admissions were included in the study. The ICU mortalities ...
Machine learning algorithms are susceptible to cyberattacks, posing security problems in computer... more Machine learning algorithms are susceptible to cyberattacks, posing security problems in computer vision, speech recognition, and recommendation systems. So far, researchers have made great strides in adopting adversarial training as a defensive strategy. Single-step adversarial training methods have been proposed as viable solutions for improving model generality and resilience. However, there has been little study to address this issue in the context of ownership-based recommendations, which may fail to capture stable results. In this work, we adapt the single-step adversarial training for ownership recommendation systems. Our main technical contributions are as follows: (1) We propose Adversarial Consumption and Production Relationship (ACPR), a model that combines factorization machine and single-step adversarial training for ownership recommendations. It enables us to take advantage of modeling consumption-production interactions with a factorization machine instead of the conventional matrix factorization method for ownership recommendations. (2) We enrich the ACPR technique with directional adversarial training and call our technique Adversarial Consumption and Production Relationship-Aware Directional Adversarial Model (ACPR-ADAM). The idea behind our ACPR-ADAM is that instead of the worst perturbation direction, the perturbation direction in the embedding space is restricted to other examples in the current embedding space, allowing us to incorporate the collaborative signal into the training process. Lastly, through extensive evaluations on Reddit and Pinterest, we demonstrate that our proposed method outperforms state-of-the-art methods. Compared with CPR and ACPR on Reddit and Pinterest datasets, our proposed ACPR-ADAM achieves 93%, 88%, and 72%, 69% enhancement in terms of AUC and HR, respectively. INDEX TERMS Adversarial attack, directional adversarial training, factorization machine, ownership recommendation.
Background and objectives: Patient–ventilator asynchronies (PVAs) are common in mechanically vent... more 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:...
2016 Computing in Cardiology Conference (CinC), 2016
Deceleration capacity (DC) of heart rate is a novel index for evaluating the activity of the auto... more Deceleration capacity (DC) of heart rate is a novel index for evaluating the activity of the autonomic nervous system (ANS). We examined whether controlling the inspiration/expiration (I/E) ratio benefits the DC analysis based on a model-generated RR interval (RRI) database. A cardiovascular system model was adopted to simulate RRI time series. The model allows analyzing the role of sympathetic and vagal activities in the ANS. The respiratory pattern can be controlled in the model. Three hundred RRI time series with random sympathetic and vagal activities were simulated. According to the ratio between the sympathetic and vagal activities (S/V ratio), these subjects were categorized into a case group (S/V>1) and a control group (S/V<1). DC was computed for each subject. The performance of DC in distinguishing the two groups was examined by the receiver operating characteristic (ROC) analysis. The respiratory period is set to 6 s. The I/E ratio was controlled as 1:2, 1:1 and 2:1...
Objective. The measurement of the static compliance of the respiratory system (C stat) during mec... more Objective. The measurement of the static compliance of the respiratory system (C stat) 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 (C qstat) 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 of C qstat for each kept cycle. Finally, the output C qstat was obtained as the average of the smallest 40 C qstat measurements. The proposed method was validated against the gold standard C stat measured from real clinical settings and compared with two reported algorithms. The gold standard C stat was 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 for C qstat versus 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 (R 2 = 0.90) between the C qstat and gold standard. Significance. The results showed that the C qstat can 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.
Machine learning classifiers are vulnerable to adversarial perturbation, and their presence raise... more Machine learning classifiers are vulnerable to adversarial perturbation, and their presence raises security concerns, especially in recommendation systems. While attacks and defense mechanisms in recommendation systems have received significant attention, Basic Iterative Method (BIM), which has been shown in Computer Vision to increase attack effectiveness by more than 60%, has received little attention in ownership recommendation. As a result, ownership recommender systems may be more sensitive to iterative perturbations, resulting in significant generalization errors. Adversarial Training, a regularization strategy that can withstand worst-case iterative perturbations, could be a viable option for improving model robustness and generalization. In this paper, we implement BIM for ownership recommendations. Through adversarial training, we propose the Adversarial Consumer and Producer Recommendation (ACPR) approach that integrates ownership features into a multi-objective pairwise ranking to capture the user’s preferences. The ACPR method learns a core embedding for each user and two transformation matrices that project the user’s core embedding into two role embeddings (i.e., a producer and consumer role) using an extension of matrix factorization. To minimize the impact of iterative perturbation, we train a consumer and producer recommender objective function using minimax adversarial training. Empirical studies on two Large-scale applications show that our method outperforms standard recommendation methods and recent methods that model ownership information.
Blood flow pulsatility is an important determinant of macro- and microvascular physiology. Pulsat... more 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.
Computer Methods and Programs in Biomedicine, 2021
BACKGROUND AND OBJECTIVE Patient-ventilator asynchrony (PVA) is the result of a mismatch between ... more 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.
Background and objective: Mismatch between invasive mechanical ventilation and the requirements o... more 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 2layer 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.
Automated detection of atrial fibrillation (AF) from electrocardiogram (ECG) recordings remains c... more Automated detection of atrial fibrillation (AF) from electrocardiogram (ECG) recordings remains challenging in real clinical settings. Deep neural networks (DNN) emerge as a promising tool for the task of AF detection. However, the success of DNN for AF detection is hampered by limited size and imbalanced number of samples in datasets. We propose a novel data augmentation strategy based on duplication, concatenation and resampling of ECG episodes to balance the number of samples among different categories as well as to increase the diversity of samples. The performance of the data augmentation method was examined on an AF database from Computing in Cardiology (CinC) challenge 2017. A 2-layer long short-term memory (LSTM) network was trained with the augmented dataset. Its ability of AF detection was evaluated using a 10-fold cross validation approach. And F1 score was adopted as the metrics. The AF detection results show that the proposed method was superior to two conventional data augmentation methods: window slicing and permutation. The network was also submitted to the evaluation system of the CinC challenge 2017. The F1 score obtained by the network using the proposed data augmentation method was close to the winner (0.82 vs. 0.83). In summary, the proposed data augmentation method provides an effective solution to enhance the dataset for improving the performance of DNN in ECG analysis. Such a method promotes the application of deep learning in the analysis of ECG, particularly when the dataset is small and imbalanced.
PWV is the speed of pulse wave propagation through the circulatory system. mPWV emerges as a nove... more 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. 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. 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 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. The...
Uploads
Papers by luping fang