Critical Network Research Articles

Multi-step policy evaluation for adaptive-critic-based tracking control towards nonlinear systems

Currently, there are a large number of tracking problems in the industry concerning nonlinear systems with unknown dynamics. In order to obtain the optimal control policy, a multi-step adaptive critic tracking control (MsACTC) algorithm is developed in this paper. By constructing a steady control law, the tracking problem is transformed into a regulation problem. The MsACTC algorithm has an adjustable convergence rate during the iterative process by incorporating a multi-step policy evaluation mechanism. The convergence proof of the algorithm is provided. In order to implement the algorithm, three neural networks are built, including the model network, the critic network, and the action network. Finally, two numerical simulation examples are given to verify the effectiveness of the algorithm. Simulation results show that the MsACTC algorithm has satisfactory performance in terms of the applicability, tracking accuracy, and convergence speed.

Weaponized disinformation spread and its impact on multi-commodity critical infrastructure networks

The well-publicized disinformation campaigns surrounding recent elections, pandemic vaccination adoption, as well as supply-chain disruptions and shortages have made historical problems of disinformation more apparent. When disinformation targets transportation infrastructure, supply chains can be disrupted, resulting in commodities shortages such as food and fuel, consequently jeopardizing the health and safety of communities. This research proposes an integrated epidemiological-optimization model that quantifies the impacts of weaponized disinformation on infrastructure networks that transport multiple commodities. The model aims to minimize the overall weighted shortage of commodities caused by different disinformation spread rates. Results show that disinformation weaponized against transportation infrastructure, potentially targeting a particular geographical region or a particular commodity, can have wide-ranging impacts across different commodities. The proposed model is applied to the multi-commodity Swedish railway network carrying 14 different categories of commodities over 1360 supply and demand nodes.

PsNRT2.3 interacts with PsNAR to promote high-affinity nitrate uptake in pea (Pisum sativum L.)

Nitrate, the primary form of nitrogen absorbed by plants, supplies essential compounds for plant growth and development. Peas are frequently used as rotation crops to improve and stabilize soil fertility. However, the determinants of nitrate uptake and transport in peas remain largely unclear, primarily due to the pea genome's complexity and size. In this study, we utilized the complete genomic information of peas to identify three PsNRT2 family genes within the pea genome. We conducted a comprehensive examination of their protein conserved domains, physicochemical properties, gene structure, and phylogenetic evolution, revealing PsNRT2.3 as the potential key gene for high-affinity nitrate transport in peas. Subcellular localization studies indicated that PsNRT2.3 resides on the plasma membrane. Using hairy root transformation, we noted the predominant expression of PsNRT2.3 in the root stele, which is inducible by nitrate. Our experiments involving overexpression and silencing methods further confirmed that PsNRT2.3 plays a key role in enhancing nitrate uptake in peas. Additionally, our work showed that PsNAR could interact with PsNRT2.3, modulating pea nitrate uptake. After silencing PsNAR, even with the normal expression of PsNRT2.3, the ability of peas to absorb nitrate was significantly reduced. In conclusion, this study identifies the high-affinity nitrate transport gene PsNRT2.3 in peas and clarifies its critical role and regulatory network in nitrate transport, contributing to a new understanding of nitrate utilization in peas.

Dynamic event-triggered-based online IRL algorithm for the decentralized control of the input and state constrained large-scale unmatched interconnected system

This article proposed a novel adaptive decentralized control (ADC) method for the continuous-time state-constrained and input-constrained large-scale unmatched interconnection system by the means of the adaptive critic design in the edge dynamic event-triggered (EDET) mechanism. The barrier function is used to transform the state-constrained system to the common system without constrained states. To overcome the influence of the unmatched interconnection terms, the auxiliary systems are devised for each subsystem of the large-scale system. Moreover, the non-quadratic utility functions are introduced to constrain the input of the auxiliary systems. The decentralized control scheme of the large-scale unmatched interconnection system can be realized by solving the optimal control schemes of the auxiliary systems. Then with the help of the single critic neural network (NN), the approximate optimal control policies are acquired by designing the approximate cost functions. After that, an integral reinforcement learning (IRL)-based scheme is proposed to solve the integral Bellman equation rather than the complex coupled Hamilton–Jacobi-Bellman equation (HJBE). Simultaneously, the EDET mechanism is introduced to reduce the computation efforts and communication loads. The asymptotic stability of the constrained large-scale unmatched interconnection system is proved. In addition, the infamous Zeno behavior is effectively avoided. Finally, two simulation cases are given to verify the effectiveness of the proposed algorithm.

Fault-tolerant tracking control for nonlinear systems with multiplicative actuator faults in view of zero-sum differential games

In view of zero-sum differential games, this paper addresses the fault-tolerant tracking control (FTTC) problem for nonlinear systems with multiplicative actuator faults. By augmenting the nominal system, the trajectory tracking problem is transformed into an optimal regulation problem. A fault observer is designed to estimate the multiplicative actuator fault factor. In order to better reflect system performance, a comprehensive performance index function including the trajectory tracking error, the control input, and the multiplicative actuator fault is constructed. By considering the control input and the multiplicative actuator fault as two players in the zero-sum differential game (ZSDG), the FTTC problem is thus transformed into a ZSDG problem. Then, a fault-tolerant control based on ZSDG is designed by solving the Hamilton–Jacobi-Isaacs (HJI) equation. Since the HJI equation is difficult to solve, a critic neural network is constructed to obtain its approximation, which constitutes the Nash equilibrium of the ZSDG. The closed-loop faulty system is guaranteed to be uniformly ultimately bounded via the Lyapunov’s direct method. The effectiveness of the developed FTTC method is demonstrated by two simulation examples.

Reinforcement learning to achieve real-time control of triple inverted pendulum

This work uses reinforcement learning (RL) to achieve the first-ever data-driven real-time control of an actual, not simulated, triple inverted pendulum (TIP) in a model-free way. A swing-up control task for the TIP is formulated as a Markov decision process with a dense reward function, then conducted in real time by using a model-free RL approach. To increase the sample efficiency of learning, a structure-aware virtual experience replay (VER) method is proposed; it works together with an off-policy actor-critic algorithm. The VER exploits the geometrically-symmetric property of TIPs to create virtual sample trajectories from measured ones, then uses the resulting multifold augmented dataset to effectively train actor and critic networks during the learning process. These structure-infused training data serve to obtain additional information and hence increase the convergence speed of network learning. We combine the proposed VER with a state-of-the-art actor-critic algorithm, and then validate its effectiveness through numerical simulations. Notably, the inclusion of VER amplifies computational efficiency, slashing the requisite trials, training steps, and overall duration by approximately 66.67%. Finally experiments demonstrate the real-time control capability of the proposed approach on an actual TIP system.

Actor Prioritized Experience Replay

A widely-studied deep reinforcement learning (RL) technique known as Prioritized Experience Replay (PER) allows agents to learn from transitions sampled with non-uniform probability proportional to their temporal-difference (TD) error. Although it has been shown that PER is one of the most crucial components for the overall performance of deep RL methods in discrete action domains, many empirical studies indicate that it considerably underperforms off-policy actor-critic algorithms. We theoretically show that actor networks cannot be effectively trained with transitions that have large TD errors. As a result, the approximate policy gradient computed under the Q-network diverges from the actual gradient computed under the optimal Q-function. Motivated by this, we introduce a novel experience replay sampling framework for actor-critic methods, which also regards issues with stability and recent findings behind the poor empirical performance of PER. The introduced algorithm suggests a new branch of improvements to PER and schedules effective and efficient training for both actor and critic networks. An extensive set of experiments verifies our theoretical findings, showing that our method outperforms competing approaches and achieves state-of-the-art results over the standard off-policy actor-critic algorithms.

Nearly optimal fault-tolerant constrained tracking for multi-axis servo system via practical terminal sliding mode and adaptive dynamic programming

In this paper, a nearly optimal tracking control is proposed for n-links robotic manipulators subject to parameter uncertainties, time-profile failures, and input saturation constraints. Firstly, the practical terminal sliding-mode (PTSM) manifold with a linear additional term is proposed to combine the system states related to joint rotation, such that the controlled states quickly fall into a tiny neighborhood of the equilibrium once they reach the PTSM manifold. Secondly, a nearly optimal sliding-mode reaching law is designed by using the adaptive dynamic programming (ADP) technique. Benefiting from a non-quadratic positive defined mapping of the proposed performance index, which relates to the derivative of the sliding-mode function, reduced-order system dynamics can be constrained to a desired region. For the bounded actuator fault caused by various inducements such as the power supply fluctuation and the wear of parts, a radial basis function neural network (RBFNN) is introduced to compensate for this, and the input saturation constraints of the controlled plant are also compensated at the same time. Innovatively, the node weights of RBFNN are updated by the critic network of the ADP framework, such that the integrity of the proposed control strategy is improved. Simulations verify the main conclusions.

Deep-Reinforcement-Learning-Based Production Scheduling in Industrial Internet of Things

The unprecedented prosperity of the Industrial Internet of Things (IIoT) promotes the traditional industry transforming into intelligent manufacturing so that the whole production process can be comprehensively controlled to achieve flexible production. Intelligent scheduling, as one of the key enabling techniques, is desired to allocate the production of several machines by an efficient solution with minimum makespan. Existing approaches adopt a fixed search paradigm based on expert knowledge to seek satisfactory solutions. However, considering the varying data distribution and large-sized of the practical problems, these methods fail to guarantee the quality of the obtained solution under the real-time requirement. To address this challenge, we formulate the production scheduling problem as a Markov decision process (MDP) and specifically design a job scheduling model made up of a job batching module for the hybrid flow-shop scheduling problem on batch processing machines (HFSP-BPM). Our proposed model consists of an actor network that learns the action under different conditions and a critic network that evaluates the action of the actor. We analyze the convergence of the model under different parameter settings to determine the optimal parameter. Extensive numerical experiments on both publicly available dataset and real steel plant production dataset demonstrate that the proposed deep reinforcement learning (DRL) approach compared with other baselines, more than 6% average improvements can be observed in many instances.

Information Security Management: American Experience

The purpose of the article is a comprehensive analysis of the American concept of threats to information security and determination of priority areas of the US’s activity in creating a secure national cyberspace. The methodological basis of the study is general scientific and special methods of cognition, in particular, systemic approach, analysis, synthesis, and logical method. Methods of content analysis, comparative and analytical monitoring of Internet resources of US government bodies responsible for information security are also used. The scientific novelty of the study consists in the expansion of ideas about theoretical aspects in the field of information security and the systematic analysis of instrumental, conceptual foundations and practical aspects of information security in the United States. Conclusions. The globalisation of information systems has created a completely new situation in the security field. In cyberspace the main threat to the US national security comes from states and intermediaries acting in their interests. They have the necessary skills and technologies to carry out destructive cyberattacks for military and political purposes, and also effectively use cyberespionage methods, which not only entails economic losses, but also causes great damage to strategically important industries for the US. The American concept covers such three key levels of cyber security as the state, private business and individual users. There are such defence priorities for the United States as ensuring the protection of critical infrastructure, information networks and systems; quality control of used IT equipment; formation of effective mechanisms of interlevel communication and raising awareness at all levels. An important component of the US National Cyber Strategy is international cooperation on information security issues. In this regard, at the international level the United States seeks to implement such opportunities as to encourage countries to increase responsibility for ensuring the security of information systems and networks at the national and global level; to create the legal regime necessary to ensure cross-border access to information; to form a regime of collective cyber defence within the framework of NATO and other bilateral and multilateral agreements with strategic partners; to preserve the maximum possible freedom of action in cyberspace in order to conduct all types of information operations both during military conflicts and in peacetime. Keywords: cyberspace, information threat, information war, information security, information management, informatisation, ICT.

Quantum Multiagent Actor–Critic Networks for Cooperative Mobile Access in Multi-UAV Systems

This paper proposes a novel algorithm, named quantum multi-agent actor-critic networks (QMACN) for autonomously constructing a robust mobile access system employing multiple unmanned aerial vehicles (UAVs). In the context of facilitating collaboration among multiple unmanned aerial vehicles (UAVs), the application of multi-agent reinforcement learning (MARL) techniques is regarded as a promising approach. These methods enable UAVs to learn collectively, optimizing their actions within a shared environment, ultimately leading to more efficient cooperative behavior. Furthermore, the principles of a quantum computing (QC) are employed in our study to enhance the training process and inference capabilities of the UAVs involved. By leveraging the unique computational advantages of quantum computing, our approach aims to boost the overall effectiveness of the UAV system. However, employing a QC introduces scalability challenges due to the near intermediate-scale quantum (NISQ) limitation associated with qubit usage. The proposed algorithm addresses this issue by implementing a quantum centralized critic, effectively mitigating the constraints imposed by NISQ limitations. Additionally, the advantages of the QMACN with performance improvements in terms of training speed and wireless service quality are verified via various data-intensive evaluations. Furthermore, this paper validates that a noise injection scheme can be used for handling environmental uncertainties in order to realize robust mobile access.

The intrinsic dimensionality of network datasets and its applications1

Modern network infrastructures are in a constant state of transformation, in large part due to the exponential growth of Internet of Things (IoT) devices. The unique properties of IoT-connected networks, such as heterogeneity and non-standardized protocol, have created critical security holes and network mismanagement. In this paper we propose a new measurement tool, Intrinsic Dimensionality (ID), to aid in analyzing and classifying network traffic. A proxy for dataset complexity, ID can be used to understand the network as a whole, aiding in tasks such as network management and provisioning. We use ID to evaluate several modern network datasets empirically. Showing that, for network and device-level data, generated using IoT methodologies, the ID of the data fits into a low dimensional representation. Additionally we explore network data complexity at the sample level using Local Intrinsic Dimensionality (LID) and propose a novel unsupervised intrusion detection technique, the Weighted Hamming LID Estimator. We show that the algortihm performs better on IoT network datasets than the Autoencoder, KNN, and Isolation Forests. Finally, we propose the use of synthetic data as an additional tool for both network data measurement as well as intrusion detection. Synthetically generated data can aid in building a more robust network dataset, while also helping in downstream tasks such as machine learning based intrusion detection models. We explore the effects of synthetic data on ID measurements, as well as its role in intrusion detection systems.

Decentralized optimal control of large-scale partially unknown nonlinear mismatched interconnected systems based on dynamic event-triggered control

In this paper, a novel decentralized control method is proposed for nonlinear mismatched large-scale interconnected systems subjected to partially unknown dynamics by designing auxiliary control for each subsystem. It is demonstrated that the control sequence consisting of the optimal control policies of auxiliary control can stabilize the system asymptotically, leading to decentralized control of the large-scale system. An integral reinforcement learning (IRL) method is firstly proposed, replacing the traditional policy iterative algorithm to analyze the optimal control problem of each edge subsystem with partially unknown dynamics. After that, the edge-based dynamic event-triggered control algorithm is proposed based on the static event-triggered control method, and an internal dynamic variable characterized by a first-order filter is defined. A single critic neural network (NN) is then designed to learn the approximate optimal control strategy under the dynamic event-triggered mechanism adaptively. The stability analysis is proposed to demonstrate that the state of the event-based pulse system is ultimately uniformly bounded (UUB) and the Zeno behavior is eliminated successfully. Finally, the effectiveness of the proposed algorithm is verified by two simulation examples to realize decentralized control of mismatched large-scale systems.

Model-free adaptive dynamic event-triggered robust control for unknown nonlinear systems using iterative neural dynamic programming

This paper proposes an adaptive dynamic event-triggered (ADET) robust control method for unknown nonlinear systems using iterative neural dynamic programming (INDP). Firstly, the ADET robust control problem is transformed into an ADET optimal control problem by introducing an infinite domain integral function. Then, dynamic variables and adaptive thresholds are designed for discrete-time systems with auxiliary variables. The proposed ADET method reduces computation and transmission costs compared to existing static and dynamic event triggering methods. INDP is utilized to learn the optimal solution of the ADET Hamilton-Jacobi-Bellman equation within the heuristic dynamic programming (HDP) framework, providing system stability and algorithm convergence. The INDP algorithm employs model, action, and critic neural networks and includes a method to directly minimize the iterative cost function in the back-propagation process. The neural network implementation of the INDP algorithm is detailed. Simulation results demonstrate that the proposed ADET method based on INDP achieves faster convergence with fewer transmitted data and control updates.

Characteristics and prognostic implications of pre-shock and normotensive cardiogenic shock in the Critical Care Cardiology Trials Network Registry

Abstract Background Limited data exist regarding the epidemiology and prognostic implications of pre-shock and normotensive cardiogenic shock (CS) states. Moreover, optimal staging of risk for pts with isolated hypotension versus isolated hypoperfusion remains unclear. Purpose We sought to compare characteristics and outcomes of pre-shock states and normotensive CS as compared with conventional CS. Methods The Critical Care Cardiology Trials Network (CCCTN) is an investigator-initiated multicenter registry of CICUs coordinated by the TIMI Study Group. Consecutive admissions (n=22,869) from 42 centers (2017-2022) were stratified into groups of isolated low CO, isolated hypotension, normotensive CS, and SCAI Stage C CS using criteria in Fig A. Results A total of 1,754 were stratified into one of the four groups: 5% isolated low CO, 9% isolated hypotension, 13% normotensive CS, and 74% SCAI stage C CS. Prior HF was more common in pts w/ isolated low CO (73%) and normotensive CS (60%), compared to pts w/ isolated hypotension (44%) or SCAI stage C CS (55%) (global p<0.001). Overall, in-hospital mortality was 18.4% with marked differences among groups (global p < 0.001; Fig B). Specifically, mortality was significantly lower in those with an isolated low CO or isolated hypotension than among the normotensive CS group (pairwise p [isolated hypotension vs. normotensive CS] = 0.03; Fig B). In contrast, mortality rates among admissions with normotensive CS vs. SCAI stage C CS did not significantly differ (pairwise p=0.08; Fig B). Conclusions Normotensive CS is associated with a >2-fold higher in-hospital mortality compared with isolated hypotension or an isolated low CO state. These findings indicate that the presence of malperfusion, irrespective of hypotension, is associated with worse outcomes in CS. These data may inform refinement of CS staging with respect to pre-shock states.Figure

MCS current practice and weaning: CCCTN survey of CICU directors at academic medical centers

Abstract Background/Introduction As the treatment of cardiogenic shock evolves, temporary mechanical circulatory support (MCS) use is increasing. Few studies have examined Cardiac Intensive Care Unit (CICU) practices for management and weaning of MCS. Purpose We aimed to describe current management and weaning practices for patients supported with an intra-aortic balloon pump (IABP) or Impella in CICUs. Methods During December 2022, a 45-question survey was electronically distributed by the Critical Care Cardiology Trials Network (CCCTN) to the CICU directors of 44 academic medical centers in North America. Descriptive statistics were used to summarize the results. Results A response rate of 84% was achieved. Among respondents, CICU bed capacity ranged from 8-32 (Mean of 17). Staffing was primarily by cardiologists without board certification in heart failure or critical care (38%). One-in-four centers reported staffing by a multi-disciplinary team consisting of cardiologists and intensivists. Most centers use heparin anticoagulation for both IABP (32, 87%) and Impella (32, 91%) and primarily used PTT to monitor anticoagulation (23, 62%) with a target of 50-70 seconds (27, 73%). While using Impella support, daily TTE (29%), plasma free hemoglobin measurements (34%) and prophylactic antibiotics (3%) were uncommon. Most institutions did not have a formal weaning protocol for IABP (78%) or Impella (66%). The majority of centers reported IABP weaning to be a subjective decision led by the CICU attending (65%), while Impella weaning was commonly a decision made by a multi-disciplinary team including the CICU attending (57%). PACs are used routinely when weaning Impella (83%) but not IABP (46%). IABP removal is most commonly done at bedside by CICU team (70%), whereas Impella removal is most commonly done in the catheterization laboratory (43%). Conclusions In this survey of academic CICUs, IABP and Impella management showed substantial variability across centers. Use of invasive hemodynamic monitoring was common but its use during weaning was not uniform. Multidisciplinary involvement was more common with Impella support and weaning when compared to IABP. These heterogeneous results highlight the need for further investigation to guide best practices in MCS management.

Sink Node Placement and Partial Connectivity in Wireless Sensor Networks.

This research delves into the aspects of communication and connectivity problems within random Wireless Sensor Networks (WSNs). It takes into account the distinctive role of the sink node, its placement, and application-specific requirements for effective communication while conserving valuable network resources. Through mathematical modeling, theoretical analysis, and simulation evaluations, we derive, compare, and contrast the probabilities of partial and full connectivity within a random WSN, factoring in network parameters and the maximum allowable hop distance/count hmax. hmax captures the diverse range of delay-sensitive requirements encountered in practical scenarios. Our research underscores the significant impact of the sink node and its placement on network connectivity and the sensor connection rate. The results exemplify a noteworthy decline in the sensor connection rate, dropping from 98.8% to 72.5%, upon relocating the sink node from the network center to the periphery. Moreover, as compared with full connectivity, partial connectivity and the sensor connection rate are more suitable metrics for assessing the communication capability of random WSNs. The results illustrate that 1.367 times more energy is required to connect less than 4% of the remote sensors, based on the examined network settings. Additionally, to increase the sensor connection rate slightly from 96% to 100%, an additional 538% more energy is required in multipath fading based on the widely adopted energy consumption model. This research and its outcomes contribute to establishing appropriate performance metrics and determining critical network parameters for the practical design and implementation of real-world wireless sensor networks.

Adaptive optimal sliding mode control for three-phase voltage source inverter: Reinforcement learning approach

The operation of the three-phase standalone inverter is affected by many factors, such as heavy changes in load, unbalanced loads, nonlinear loads, system uncertainties and external disturbances. These are critical weaknesses for the control system of the three-phase inverter to reach robust optimal performance. This paper proposes an adaptive optimal sliding mode control (AOSMC) scheme for a three-phase nonlinear uncertain inverter. This AOSMC strategy solves the problems of nonlinear optimization by adaptive dynamic programming, one of the techniques of reinforcement learning, and overcomes the uncertainties and disturbance effects by a disturbance observer–based sliding mode controller. This algorithm uses only one neural network to approximate the critic; therefore, the burden of computation is significantly reduced. Both the weight matrix of the critic network and the disturbance observer are asymptotically stable. The overall system is guaranteed to be ultimately uniformly bounded stable via Lyapunov stable theory. The simulation is conducted to validate the insensitivity of the proposed AOSMC algorithm to working conditions. Also, the competitive results are presented to demonstrate the improvement of the proposed AOSMC scheme in comparison to some other existing controllers.

Abstract 11821: Association of Hemoglobin With Mortality Outcomes in the Critical Care Cardiology Trials Network (CCCTN) Registry

Introduction: Anemia is common in critically ill patients and is a known risk indicator for poor outcomes. Little is known about the impact of anemia on short term outcomes in the setting of cardiac critical illness. We examined the association of varying degrees of anemia with in-hospital outcomes in cardiac intensive care unit (CICU) admissions. Methods: The Critical Care Cardiology Trials Network (CCCTN) is a multicenter network of tertiary CICUs in North America coordinated by the TIMI Study Group. During annual campaigns, each center contributes at least 2-mo of all consecutive medical CICU admissions. Admissions were stratified according to baseline hemoglobin (Hgb). The association of Hgb and in-hospital mortality, CICU length of stay and hospital length of stay was examined by logistic regression with multivariable adjustment. Results: Among 17,759 admissions, the median baseline Hgb was 12.0 g/dL (IQR 10.0, 13.8). Patients with lower Hgb tended to be older and female with lower body weight and a greater burden of comorbidities. After adjusting for sex and other relevant covariates, a lower baseline Hgb was associated with higher odds of in-hospital death (p-trend <0.0001), and longer CICU and hospital length of stay (p-trend <0.0001). Conclusions: Anemia is prevalent in half of patients admitted to the CICU. Lower Hgb level, even well above standard transfusion thresholds, is independently associated with higher in-hospital mortality and longer length of stay in cardiac critical illness.

Abstract 13676: Characteristics and Outcomes of Patients With Clinically Significant Valvular Heart Disease and Cardiogenic Shock Admitted to Cardiac Intensive Care Units

Introduction: Little is known about the characteristics & outcomes of patients (pts) with significant valvular heart disease (VHD) and cardiogenic shock (CS). In addition, differences between pts with CS due to VHD vs VHD as “bystander” to CS have not been described. Methods: The Critical Care Cardiology Trials Network is a multicenter network of CICUs in North America coordinated by the TIMI Study Group. Participating centers capture all CICU admissions for 2 months annually. Pts with CS and significant VHD were classified as having CS due to VHD or CS with concomitant VHD not felt to be the cause of CS. Admissions from 2017 - 2021 were analyzed. Results: Of 3,974 CICU admissions with CS, 165 (4.2%) had CS due to VHD and 769 (19.4%) had CS with concomitant VHD. Pts with CS due to VHD were older (71 vs 67 y), less commonly had a hx of HF (47.3% vs 73.1%), & more commonly had LVEF>50% on admission (51.5% vs 16.8%); p<0.001 for all. Resource use varied significantly (Fig). Nearly 1/3 of pts (30.3%) with CS due to VHD underwent surgical/transcatheter valve procedures during CICU admission, contrasting with those with CS and concomitant VHD (13.5%; p<0.001). In-hospital mortality was similar between the 2 VHD groups (39.4% vs 33.8%, p=0.17). In-hospital mortality for allcomers with CS and VHD (concomitant or causative) was higher than in those with CS and no VHD (34.8% vs 29.7%; p=0.003), including adjusted for age, sex, SOFA score, and pre-CICU cardiac arrest (aOR 1.23; 1.03-1.47). Mortality was numerically lower in pts undergoing a valve procedure (23.4% died; 16% with CS due to VHD, 26.9% with CS and concomitant VHD; p=0.13), compared to pts who did not (37.1% overall; 49.6% & 34.9%, respectively; p=0.003). Conclusions: Although VHD is the primary cause of CS in a minority of pts, the presence of significant VHD is associated with higher mortality. In contemporary CICUs, procedural intervention is undertaken in ~1/3 of cases and mortality was numerically lower in pts who underwent valve procedures.