TransFOL: A Logical Query Model for Complex Relational Reasoning in Drug-Drug Interaction.

Predicting drug-drug interaction (DDI) plays a crucial role in drug recommendation and discovery. However, wet lab methods are prohibitively expensive and time-consuming due to drug interactions. In recent years, deep learning methods have gained widespread use in drug reasoning. Although these methods have demonstrated effectiveness, they can only predict the interaction between a drug pair and do not contain any other information. However, DDI is greatly affected by various other biomedical factors (such as the dose of the drug). As a result, it is challenging to apply them to more complex and meaningful reasoning tasks. Therefore, this study regards DDI as a link prediction problem on knowledge graphs and proposes a DDI prediction model based on Cross-Transformer and Graph Convolutional Networks (GCN) in first-order logical query form, TransFOL. In the model, a biomedical query graph is first built to learn the embedding representation. Subsequently, an enhancement module is designed to aggregate the semantics of entities and relations. Cross-Transformer is used for encoding to obtain semantic information between nodes, and GCN is used to gather neighbour information further and predict inference results. To evaluate the performance of TransFOL on common DDI tasks, we conduct experiments on two benchmark datasets. The experimental results indicate that our model outperforms state-of-the-art methods on traditional DDI tasks. Additionally, we introduce different biomedical information in the other two experiments to make the settings more realistic. Experimental results verify the strong drug reasoning ability and generalization of TransFOL in complex settings. Data and code are available at https://github.com/Cheng0829/TransFOL.

Knowledge enhanced attention aggregation network for medicine recommendation.

The combination of deep learning and the medical field has recently achieved great success, particularly in recommending medicine for patients. However, patients' clinical records often contain repeated medical information that can significantly impact their health condition. Most existing methods for modeling longitudinal patient information overlook the impact of individual diagnoses and procedures on the patient's health, resulting in insufficient patient representation and limited accuracy of medicine recommendations. Therefore, we propose a medicine recommendation model called KEAN, which is based on an attention aggregation network and enhanced graph convolution. Specifically, KEAN can aggregate individual diagnoses and procedures in patient visits to capture significant features that affect patients' diseases. We further incorporate medicine knowledge from complex medicine combinations, reduce drug-drug interactions (DDIs), and recommend medicines that are beneficial to patients' health. The experimental results on the MIMIC-III dataset demonstrate that our model outperforms existing advanced methods, which highlights the effectiveness of the proposed method.

From Simple to Complex Scenes: Learning Robust Feature Representations for Accurate Human Parsing.

Human parsing has attracted considerable research interest due to its broad potential applications in the computer vision community. In this paper, we explore several useful properties, including high-resolution representation, auxiliary guidance, and model robustness, which collectively contribute to a novel method for accurate human parsing in both simple and complex scenes. Starting from simple scenes: we propose the boundary-aware hybrid resolution network (BHRN), an advanced human parsing network. BHRN utilizes deconvolutional layers and multi-scale supervision to generate rich high-resolution representations. Additionally, it includes an edge perceiving branch designed to enhance the fineness of part boundaries. Building on BHRN, we construct a dual-task mutual learning (DTML) framework. It not only provides implicit guidance to assist the parser by incorporating boundary features, but also explicitly maintains the high-order consistency between the parsing prediction and the ground truth. Toward complex scenes: we develop a domain transform method to enhance the model robustness. By transforming the input space from the spatial domain to the polar harmonic Fourier moment domain, the mapping relationship to the output semantic space is highly stable. This transformation yields robust representations for both clean and corrupted data. When evaluated on standard benchmark datasets, our method achieves superior performance compared to state-of-the-art human parsing methods. Furthermore, our domain transform strategy significantly improves the robustness of DTML dramatically in most complex scenes.

High-Performance All-Textile Triboelectric Nanogenerator toward Intelligent Sports Sensing and Biomechanical Energy Harvesting.

Merging textiles with advanced energy harvesting technology via triboelectric effects brings novel insights into self-powered wearable textile electronics. However, fabrication of a comfortable textile-based triboelectric nanogenerator (TENG) with high outputs remains challenging. Herein, we propose a highly flexible, tailorable, single-electrode all-textile TENG (t-TENG) with both wear comfort and high outputs. A dielectric modulated porous composite coating containing poly(vinylidene fluoride)-hexafluoropropylene copolymer and barium titanate nanoparticles is constructed on conductive fabric to counterpart with highly positive glass fiber fabric through knotted yarn bonding, maintaining the superiority of textiles and strong triboelectricity. Through the synergistic optimization of charge storage via dielectric modulation and charge dissipation offset by electrical poling, remarkable outputs (261 V, 1.5 µA, and 12.7 nC) are obtained from a miniaturized, lightweight t-TENG (2 × 2 cm2, 130 mg) with an instantaneous power density of 654.48 mW·m-2, as well as excellent electrical robustness and device durability over 20,000 cycles. The t-TENG also exhibits a high sensitivity of 3.438 V·kPa-1 in the force region (1-10 N), demonstrating great potential in TENG-based intelligent sports sensing applications for monitoring and correcting the basketball shooting hand and foot arch posture. Furthermore, over 110 light-emitting diode arrays can be lightened up by gently tapping this miniaturized t-TENG. It also offers a wearable power source scheme through integrating the single-electrode device into clothing and utilizing the skin as the grounded electrode, revealing its ease of integration and biomechanical energy harvesting capability. This work provides an attractive paradigm for next-generation textile electronics with well-balanced device performance and wear comfort.

SADR: Self-Supervised Graph Learning With Adaptive Denoising for Drug Repositioning.

Traditional drug development is often high-risk and time-consuming. A promising alternative is to reuse or relocate approved drugs. Recently, some methods based on graph representation learning have started to be used for drug repositioning. These models learn the low dimensional embeddings of drug and disease nodes from the drug-disease interaction network to predict the potential association between drugs and diseases. However, these methods have strict requirements for the dataset, and if the dataset is sparse, the performance of these methods will be severely affected. At the same time, these methods have poor robustness to noise in the dataset. In response to the above challenges, we propose a drug repositioning model based on self-supervised graph learning with adptive denoising, called SADR. SADR uses data augmentation and contrastive learning strategies to learn feature representations of nodes, which can effectively solve the problems caused by sparse datasets. SADR includes an adaptive denoising training (ADT) component that can effectively identify noisy data during the training process and remove the impact of noise on the model. We have conducted comprehensive experiments on three datasets and have achieved better prediction accuracy compared to multiple baseline models. At the same time, we propose the top 10 new predictive approved drugs for treating two diseases. This demonstrates the ability of our model to identify potential drug candidates for disease indications.

Investigation of the predictive value of a novel algorithm based on coronary CT angiography regarding fractional flow reserve and revascularization in patients with stable coronary artery disease.

Fractional flow reserve (FFR) has been established as a gold standard for functional coronary ischemia. At present, the FFR can be calculated from coronary computed tomography angiography (CCTA) images (CT-FFR). Previous studies have suggested that CT-FFR outperforms CCTA and invasive coronary angiography (ICA) in determining hemodynamic significance of stenoses. Recently, a novel automatical algorithm of CT-FFR called RuiXin-FFR has been developed. The present study is designed to investigate the predictive value of this algorithm and its value in therapeutic decision making. The present study retrospectively included 166 patients with stable coronary artery disease (CAD) who underwent CCTA screening and diagnostic ICA examination at Peking University People's Hospital, in 73 of whom wire-derived FFR was also measured. CT-FFR analyses were performed with a dedicated software. All patients were followed up for at least 1 year. We validated the accuracy of RuiXin-FFR with invasive FFR as the standard of reference, and investigated the role of RuiXin-FFR in predicting treatment strategy and long-term prognosis. The mean age of the patients was 63.3 years with 63.9% male. The CT-FFR showed a moderate correlation with wire-derived FFR (r = 0.542, p < 0.0001) and diagnostic accuracy of 87.6% to predict myocardial ischemia (AUC: 0.839, 95% CI 0.728-0.950), which was significantly higher than CCTA and ICA. In the multivariate logistic regression analysis, CT-FFR ≤ 0.80 was an independent predictor of undergoing coronary revascularization (OR: 45.54, 95% CI 12.03-172.38, p < 0.0001), whereas CT-FFR > 0.80 was an independent predictor of non-obstructive CAD (OR: 14.67, 95% CI 5.42-39.72, p < 0.0001). Reserving ICA and revascularization for vessels with positive CT-FFR could have reduced the rate of ICA by 29.6%, lowered the rate of ICA in vessels without stenosis > 50% by 11.7%, and increased the rate of revascularization in patients receiving ICA by 21.2%. The average follow-up was 23.7 months, and major adverse cardiovascular events (MACE) occurred in 11 patients. The rate of MACE was significantly lower in patients with CT-FFR > 0.80. The new algorithm of CT-FFR can be used to predict the invasive FFR. The RuiXin-FFR can also provide useful information for the screening of patients in whom further ICA is indeed needed and prognosis evaluation.

A study on fiber Bragg grating (FBG) strain sensor based on symmetrical lever structure.

Thin-walled structure deformation detection technology is one of the key technologies for structural health monitoring and fault diagnosis of high-end mechanical equipment. Aiming at the problem that the existing Fiber Bragg Grating (FBG) strain sensor is difficult to effectively measure the deformation of thin-walled structures, an FBG strain sensor based on a symmetrical lever structure is proposed. The sensitivity of the sensor is analyzed theoretically, and the sensor is simulated and analyzed by the SOLIDWORKS and Abaqus software, and then, the structural parameters are optimized. According to the simulation results, the sensor is developed and a strain testing system is set up to test the performance of the sensor. The results indicate that the sensor sensitivity is ∼6.6 pm/µÎµ, which is about 5.5 times that of bare FBG. Its strain measurement sensitivity and stability are much higher than those of bare FBG, thus meeting the strain detection requirements of thin-walled structural parts during deformation. Moreover, the linearity is more than 99%, which enables the accurate measurement of tiny strains caused by the deformation and reconstruction of the thin-walled structure by the strain sensor. The results of this study provide a reference for the development of like sensors and a further improvement in the sensitivity of the optic-fiber strain sensor.

Semantic-enhanced Graph Contrastive Learning with Adaptive Denoising for Drug Repositioning.

The traditional drug development process requires a significant investment in workforce and financial resources. Drug repositioning as an efficient alternative has attracted much attention during the last few years. Despite the wide application and success of the method, there are still many shortcomings in the existing model. For example, sparse datasets will seriously affect the existing methods' performance. Additionally, these methods do not pay attention to the noise in datasets. In response to the above defects, we propose a semantic-enriched augmented graph contrastive learning with an adaptive denoising method, called SGCD. This method enhances data from the perspective of the embedding layer, deeply mines potential neighborhood relation-ships in semantic space, and combines similar drugs in the semantic neighborhoods into prototype comparison targets, thus effectively mitigating the impact of data sparsity on the model. Moreover, to enhance the model's robustness to noisy data, we use the adaptive denoising method, which can effectively identify noisy data in the training process. Exhaustive experiments on multiple real datasets show the effectiveness of the proposed model. The code implementation is available at https://github.com/yuhuimin11/SGCD-master.

Automatic International Classification of Diseases Coding via Note-Code Interaction Network with Denoising Mechanism.

Clinical notes are comprehensive files containing explicit information about a patient's visit. However, accurately assigning medical codes from clinical documents can be a persistent challenge due to the complexity of clinical data and the vast range of medical codes. Moreover, the large volume of medical records, the noisy medical records, and the uneven quality of coders all negatively impact the quality of the final codes. Deep learning technology has recently been integrated into automatic International Classification of Diseases (ICD) coding tasks to improve accuracy. Nevertheless, the imbalanced class problem, the complexness of code associations, and the noise in lengthy records still restrict the advancement of ICD coding tasks in deep learning. Thus, we present the Note-code Interaction Denoising Network (NIDN) that employs the self-attention mechanism to pull critical semantic features in electronic medical records (EMRs). Our model utilizes the label attention mechanism for retaining code-specific text expression. We introduce Clinical Classifications Software coding for multitask learning, capturing the functional relationships of medical coding to oblige in model prediction. To minimize the impact of noise on model prediction and improve the label distribution imbalance, a denoising module is introduced to filter noise. Our practical consequences indicate that the model NIDN exceeds competitive models on a third version of Medical Information Mart for Intensive Care data set.

Self-Assembly and Photothermal Conversion of MetallaRussian Doll and Metalla[2]catenanes Induced via Multiple Stacking Interactions.

A variety of organometallic supramolecular architectures have been constructed over the past decades and their properties were also explored via different strategies. However, the synthesis of metalla-Russian doll is still a fascinating challenge. Herein, a series of new coordination supramolecular complexes, including a metalla-Russian doll, metalla[2]catenanes, and metallarectangles, were synthesized by using meticulously selected Cp*Rh (Cp* = η5-C5Me5) building units (E1, E2, and E3) and three rigid anthracylpyridine ligands (L1, L2, and L3) via a self-assembly strategy. While the combination of the short ligand L1 and E1 or E2 generated two metallarectangles, the longer ligand L2 containing an alkynyl group resulted in two new [2]catenanes, most likely due to which the strong electron-donating effect of alkynyl groups causes self-accumulation. Interestingly, an unusual Russian doll assembly was obtained through the reaction of L3 and E3 based on sextuple π···π stacking interactions. Furthermore, the dynamic structural conversion between [2]catenanes and the corresponding metallarectangles could be observed through concentration-, solvent-, and guest-induced effects. The [2]catenane complexes 4b displayed efficient photothermal conversion efficiency in solution (20.2%), in comparison with other organometallic macrocycles. We believe that π···π stacking interactions generate active nonradiative pathways and promote radiative photodeactivation pathways. This study proves the versatility of half-sandwich building units, not only to build complicated supramolecular topologies but also in effective functional materials for various appealing applications.

DGCL: Distance-wise and Graph Contrastive Learning for medication recommendation.

Medicine recommendation aims to provide a combination of medicine based on the patient's electronic health record (EHR), which is an essential task in healthcare. Existing methods either base recommendations on EHRs or provide models with knowledge of drug-drug interactions (DDIs) to achieve DDI reduction. However, the former models the patient's health history but ignores undesirable DDIs, while the latter lacks mining of patient health records and gets low recommendation accuracy. Therefore, this study contributes to research on personalized medication recommendations that consider drug interaction effects and models the patient's past medical history. In this paper, the Distance-wise and Graph Contrastive Learning (DGCL) framework is proposed. Specifically, we develop a two-stage neural network module for clinical record learning. We propose the distance detection loss to model the difference between the output distribution of current cases and historical records. In the DDI recognition and control task, DGCL proposes a graph contrastive learning method to jointly train the DDI knowledge graph and the electronic record graph, thereby effectively controlling the level of DDI for recommended medications. By comparing the performance on the MIMIC-III dataset with several baselines, DGCL outperforms other models in terms of efficacy and safety.

A novel optical coherence tomography­based calcium scoring system can predict the stent expansion of moderate and severe calcified lesions.

Coronary calcified lesions can exert serious effects on stent expansion. A calcium scoring system, based on optical coherence tomography (OCT), has been previously developed to identify relatively mild calcified lesions that would benefit from plaque modification procedures. Therefore, the present study aimed to establish a novel OCT-based scoring system to predict the stent expansion of moderate and severe calcified lesions. A total of 33 patients who underwent percutaneous coronary intervention (PCI; 34 calcified lesions were observed using coronary angiography) were retrospectively included in the present study. Coronary angiography and OCT images were subsequently reviewed and analyzed. Furthermore, a calcium scoring system was developed based on the results of multivariate analysis before the optimal threshold for the prediction of stent underexpansion in patients with moderate and severe calcified lesions was determined. The mean age of the patients was 67±10 years. The present analysis demonstrated that the final post-PCI median stent expansion was 70.74%, where stent underexpansion (defined as stent expansion <80%) was observed in 23 lesions. The mean maximum calcium arc, length and thickness, which were assessed using OCT, were found to be 230˚, 25.10 mm and 1.18 mm, respectively. A multivariate logistic regression model demonstrated that age and the maximum calcium arc were independent predictors of stent underexpansion. A novel calcium scoring system was thereafter established using the following formula: (0.16 x age) + (0.03 x maximum calcium arc) according to the ß-coefficients in the multivariate analysis, with the optimal cut-off value for the prediction of stent underexpansion being 16.87. Receiver operating characteristic curve analysis demonstrated that this novel scoring system yielded a larger area under the curve value compared with that from a previous study's scoring system. Therefore, in conclusion, since the calcium scoring system of the present study based on age and the maximum calcium arc obtained from OCT was specifically developed in the subjects with moderate and severe calcified lesions, it may be more accurate in predicting the risk of stent underexpansion in these patients.

Femtosecond UV Laser Ablation Characteristics of Polymers Used as the Matrix of Astronautic Composite Material.

Ultrafast laser processing has recently emerged as a new tool for processing fiber-reinforced polymer (FRP) composites. In the astronautic industry, the modified epoxy resin (named 4211) and the modified cyanate ester resin (known as BS-4) are two of the most widely used polymers for polymer-based composites. To study the removal mechanism and ablation process of different material components during the ultrafast laser processing of FRPs, we isolated the role of the two important polymers from their composites by studying their femtosecond UV laser (260 fs, 343 nm) ablation characteristics for controllable machining and understanding the related mechanisms. Intrinsic properties for the materials' transmission spectrum, the absorption coefficient and the optical bandgap (Eg), were measured, derived, and compared. Key parameters for controllable laser processing, including the ablation threshold (Fth), energy penetration depth (δeff), and absorbed energy density (Eabs) at the ablation threshold, as well as their respective "incubation" effect under multiple pulse excitations, were deduced analytically. The ablation thresholds for the two resins, derived from both the diameter-regression and depth-regression techniques, were compared between resins and between techniques. An optical bandgap of 3.1 eV and 2.8 eV for the 4211 and BS-4 resins, respectively, were obtained. A detectable but insignificant-to-ablation difference in intrinsic properties and ablation characteristics between the two resins was found. A systematic discrepancy, by a factor of 30~50%, between the two techniques for deriving ablation thresholds was shown and discussed. For the 4211 resin ablated by a single UV laser pulse, a Fth of 0.42 J/cm2, a δeff of 219 nm, and an Eabs of 18.4 kJ/cm3 was suggested, and they are 0.45 J/cm2, 183 nm, and 23.2 kJ/cm3, respectively, for the BS-4 resin. The study may shed light on the materials' UV laser processing, further the theoretical modeling of ultrafast laser ablation, and provide a reference for the femtosecond UV laser processing characteristics of FRPs for the future.

Blind-noise image denoising with block-matching domain transformation filtering and improved guided filtering.

The adaptive block size processing method in different image areas makes block-matching and 3D-filtering (BM3D) have a very good image denoising effect. Based on these observation, in this paper, we improve BM3D in three aspects: adaptive noise variance estimation, domain transformation filtering and nonlinear filtering. First, we improve the noise-variance estimation method of principle component analysis using multilayer wavelet decomposition. Second, we propose compressive sensing based Gaussian sequence Hartley domain transform filtering to reduce noise. Finally, we perform edge-preserving smoothing on the preprocessed image using the guided filtering based on total variation. Experimental results show that the proposed denoising method can be competitive with many representative denoising methods on the evaluation criteria of PSNR. However, it is worth further research on the visual quality of denoised images.

Left ventricular function and coronary microcirculation in patients with mild reduced ejection fraction after STEMI.

BACKGROUND: The characteristics of heart failure (HF) with mildly reduced ejection fraction (EF) (HFmrEF) overlap with those of HF with reduced EF (HFrEF) and HF with preserved EF (HFpEF) and need to be further explored. This study aimed to evaluate left ventricular (LV) function and coronary microcirculation in patients with mildly reduced ejection fraction after acute ST-segment elevation myocardial infarction (STEMI). METHODS: We enrolled 119 patients with STEMI who had undergone speckle tracking imaging and myocardial contrast echocardiography during hospitalization from June 2016 to June 2021. They were classified into normal, HFmrEF, and HFrEF groups according to their left ventricular EF (LVEF): ≥ 50%, 40-50%, and ≤ 40%, respectively. The data of the HFmrEF group were analyzed and compared with those of the normal and HFrEF groups. RESULTS: HFmrEF was observed in 32 patients (26.9%), HFrEF in 17 (14.3%), and normal LVEF in 70 patients (58.8%). The mean global longitudinal strain (GLS) of all patients was - 11.9 ± 3.8%. The GLS of HFmrEF patients was not significantly different from that of the HFrEF group (- 9.9 ± 2.5% and - 8.0 ± 2.3%, respectively, P = 0.052), but they were both lower than that of the normal group (- 13.8% ± 3.5%, P < 0.001). The HFmrEF group exhibited significantly poorer myocardial perfusion index (1.24 ± 0.33) than the normal group (1.08 ± 0.14, P = 0.005) but displayed no significant difference from the HFrEF group (1.18 ± 0.19, P = 0.486). Moreover, a significant difference in the incidence of regional wall motion (WM) abnormalities in the three groups was observed (P = 0.009), and the WM score index of patients with HFmrEF was 1.76 ± 0.30, similar to that of patients with HFrEF (1.81 ± 0.43, P = 0.618), but poorer than that in the normal group (1.33 ± 0.25, P < 0.001). CONCLUSIONS: GLS is a more sensitive tool than LVEF for detecting LV systolic dysfunction. The LV systolic function, coronary microcirculation, and WM in patients with HFmrEF was poorer than that of patients with normal LVEF, but comparable to that in patients with HFrEF. Patients with HFmrEF after STEMI require more attention and appropriate management.

Early Prediction of Cardiogenic Shock Using Machine Learning.

Cardiogenic shock (CS) is a severe condition with in-hospital mortality of up to 50%. Patients who develop CS may have previous cardiac history, but that may not always be the case, adding to the challenges in optimally identifying and managing these patients. Patients may present to a medical facility with CS or develop CS while in the emergency department (ED), in a general inpatient ward (WARD) or in the critical care unit (CC). While different clinical pathways for management exist once CS is recognized, there are challenges in identifying the patients in a timely manner, in all settings, in a timeframe that will allow proper management. We therefore developed and evaluated retrospectively a machine learning model based on the XGBoost (XGB) algorithm which runs automatically on patient data from the electronic health record (EHR). The algorithm was trained on 8 years of de-identified data (from 2010 to 2017) collected from a large regional healthcare system. The input variables include demographics, vital signs, laboratory values, some orders, and specific pre-existing diagnoses. The model was designed to make predictions 2 h prior to the need of first CS intervention (inotrope, vasopressor, or mechanical circulatory support). The algorithm achieves an overall area under curve (AUC) of 0.87 (0.81 in CC, 0.84 in ED, 0.97 in WARD), which is considered useful for clinical use. The algorithm can be refined based on specific elements defining patient subpopulations, for example presence of acute myocardial infarction (AMI) or congestive heart failure (CHF), further increasing its precision when a patient has these conditions. The top-contributing risk factors learned by the model are consistent with existing clinical findings. Our conclusion is that a useful machine learning model can be used to predict the development of CS. This manuscript describes the main steps of the development process and our results.

The Coronary Angiography-Derived Index of Microcirculatory Resistance Predicts Left Ventricular Performance Recovery in Patients with ST-Segment Elevation Myocardial Infarction.

Objectives: The present study is designed to investigate the impact of coronary angiography-derived index of microcirculatory resistance (caIMR) on left ventricular performance recovery. Background: IMR has been established as a gold standard for coronary microvascular assessment and a predictor of left ventricular recovery after ST-segment elevation myocardial infarction (STEMI). CaIMR is a novel and accurate alternative of IMR. Methods: The present study retrospectively included 80 patients with STEMI who underwent primary percutaneous coronary intervention (PCI). We offline performed the post-PCI caIMR analysis of the culprit vessel. Echocardiography was performed within the first 24 hours and at 3 months after the index procedure. Left ventricular recovery was defined as the change in left ventricular ejection fraction (LVEF) more than zero. Results: The mean age of the patients was 58.0 years with 80.0% male. The average post-PCI caIMR was 43.2. Overall left ventricular recovery was seen in 41 patients. Post-PCI caIMR (OR: 0.948, 95% CI: 0.916-0.981, p = 0.002), left anterior descending as the culprit vessel (OR: 3.605, 95% CI: 1.23-10.567, p = 0.019), and male (OR: 0.254, 95% CI: 0.066-0.979, p = 0.047) were independent predictors of left ventricular recovery at 3 months follow-up. A predictive model was established with the best cutoff value for the prediction of left ventricular recovery 2.33 (sensitivity 0.610, specificity 0.897, and area under the curve 0.765). In patients with a predictive model score less than 2.33, the LVEF increased significantly at 3 months. Conclusions: The post-PCI caIMR can accurately predict left ventricular functional recovery at 3 months follow-up in patients with STEMI treated by primary PCI, supporting its use in clinical practice.

Titration of Ventilator Settings to Target Driving Pressure and Mechanical Power.

PURPOSE: Driving pressure (ΔP) and mechanical power (MP) may be important mediators of lung injury in acute respiratory distress syndrome (ARDS) however there is little evidence for strategies directed at lowering these parameters. We applied predictive modeling to estimate the effects of modifying ventilator parameters on ΔP and MP. METHODS: 2,622 ARDS patients (Berlin criteria) from the Medical Information Mart for Intensive Care IV database (MIMIC-IV version1.0) admitted to the intensive care unit (ICU) at Beth Israel Deaconess Medical Center between 2008 and 2019 were included. Flexible confounding-adjusted regression models for time varying data were fit to estimate the effects of adjusting PEEP and tidal volume (VT) on ΔP, and adjusting VT and respiratory rate (f) on MP. RESULTS: Reduction in VT reduced ΔP and MP, with more pronounced effect on MP with lower compliance. Strategies reducing f, consistently increased MP (when VT was adjusted to maintain consistent minute ventilation). Adjustment of PEEP yielded a U-shaped effect on ΔP. CONCLUSIONS: This novel conditional modeling confirmed expected response patterns for ΔP, with the response to adjustments depending on patients' lung mechanics. Furthermore a VT -driven approach should be favored over a f -driven approach when aiming to reduce MP.

On the spurious resonance modes of time domain integral equations for analyzing acoustic scattering from penetrable objects.

The interior resonance problem of time domain integral equations (TDIEs) formulated to analyze acoustic field interactions on penetrable objects is investigated. Two types of TDIEs are considered: The first equation, which is termed the time domain potential integral equation (TDPIE), suffers from the interior resonance problem, i.e., its solution is replete with spurious modes that are excited at the resonance frequencies of the acoustic cavity in the shape of the scatterer. Numerical experiments demonstrate that, unlike the frequency-domain integral equations, the amplitude of these modes in the time domain could be suppressed to a level that does not significantly affect the solution. This is achieved by increasing the numerical solution accuracy through the use of a higher-order discretization in space and the band limited approximate prolate spheroidal wave function with high interpolation accuracy as basis function in time. The second equation is obtained by linearly combining TDPIE with its normal derivative. The solution of this equation, which is termed the time domain combined potential integral equation (TDCPIE), does not involve any spurious interior resonance modes but it is not as accurate as the TDPIE solution at non-resonance frequencies. In addition, TDCPIE's discretization calls for treatment of hypersingular integrals.