Search

Very fast imaging techniques can enhance the precision of image-guided radiation therapy, which can be useful for external beam radiation therapy. This work aims to develop a deep learning (DL)-based image-guide framework to enable fast volumetric image reconstruction for accurate target localization for treating lung cancer patients with gating, and it is presented in the context of FLASH which leverages ultrahigh dose-rate radiation to enhance the sparing of organs at risk without compromising tumor control probability. The proposed framework comprises four modules, including orthogonal kV X-ray projection acquisition, DL-based volumetric image generation, image quality analyses, and proton water equivalent thickness (WET) evaluation. We investigated volumetric image reconstruction using kV projection pairs with four different source angles. Thirty patients with lung targets were identified from an institutional database, each patient having a 4-D computed tomography (CT) dataset with ten respiratory phases. Considering all evaluation metrics, the kV projections with source angles of 135° and 225° yielded the optimal volumetric images. The patient-averaged mean absolute error, peak signal-to-noise ratio, structural similarity index measure, and WET error were $75\pm 22$ hounsfield unit, $19\pm 3$ .7 dB, $0.938\pm 0.044$ , and −1.3%±4.1%. The proposed framework can rapidly deliver volumetric images to potentially guide proton FLASH treatment delivery systems.

This article investigates the synthesis of distributed attackers for distributed discrete-event systems, where the distributed attackers could implement sensor insertion/deletion/replacement attacks and actuator enablement/disablement attacks. The distributed attackers to be synthesized should fulfill the following goals: 1) Each local sensor–actuator attacker needs to remain covert against all the distributed monitors, which are used to detect the existence of attacks; and 2) the distributed sensor–actuator attackers need to cooperate to implement attacks such that the damage infliction could be realized. To solve this distributed attacker synthesis problem, our solution methodology is to first model the original problem as a distributed supervisor synthesis problem by modeling each system component as a finite-state automaton and then perform incremental synthesis with a heuristic that takes the special structure of this distributed synthesis problem into consideration. The effectiveness of the proposed approach is illustrated via an example regarding the transportation task of multiple robots.

Brandin and Wonham introduced a timed supervisory control framework in 1994. The key innovations arguably were the introduction of tick events to model the passing of time, and the introduction of event forcing. The framework enriches the Ramadge-Wonham framework, albeit with the cost to deal with high modeling and synthesis complexities due to explicit enumeration of ticks. To address these challenges, the framework presented forgoes the explicit enumeration of tick events and instead relies on time-interval automata whose transitions are labeled by events equipped with discrete time-intervals. In contrast to the time-interval resetting mechanism of the Brandin-Wonham framework, where the time-interval of each event is reset either straight after the event occurs or when the event is not defined at a destination state, in the present framework, the time-interval of each event is reset after a new state is reached, where this event is defined. This new framework contains the Brandin-Wonham framework as a special case. After introducing the plant modeled as a free (time-interval) language, and considering the impact of event forcing on plant behaviors with the concept of time-coerced languages, the supervisory control problem is introduced, whose requirements also are time-interval languages. After presenting the concept of controllability in the time-interval framework and demonstrating the surprising existence of supremal time-coerced controllable sublanguages for a given plant and requirements, even though time-coerced languages are typically not closed under union in contrast to free languages, a new supremal controllable sublanguage synthesis algorithm is presented.

This work investigates the existence of fortified supervisors against covert actuator attackers. For a supervisor $S$ that is nonresilient against covert actuator attackers, a fortified supervisor $S^{\prime }$ satisfies two requirements: 1) $S^{\prime }$ is resilient against any covert actuator attacker, and 2) the original closed-behavior of the closed-loop system under $S$ is preserved. We design a sound and complete procedure to show the problem of determining the existence of a fortified supervisor against covert actuator attackers is decidable. The decidability result is also extended to the case against the worst-case attacker.

This paper presents an overview of the networked supervisory control frameworks for discrete event systems with imperfect communication networks, which are divided into the centralized setup, the decentralized setup and the distributed setup. The state-of-the-art works on networked supervisory control of discrete-event systems addressing the channel imperfections that are caused by channel delays, data losses or non-FIFO channels are discussed. By presenting the key concepts and main results of each representative work, we analyze the pros and cons of different approaches. Finally, we also provide a summary of the existing works, which roughly follow two different lines of thinking and result in two different verification or synthesis approaches. The first approach utilizes simple, non-networked plant models but relies on the development of sophisticated concepts of network controllability and observability to capture network imperfections, while the second approach embeds relatively complex yet verifiable channel models into the model of the networked plant and adopts the standard concepts of controllability and observability for the (verification and) synthesis of networked supervisors. Some future research topics are also presented.

In this article, we consider the problem of synthesizing the supremal covert damage-reachable attacker, in the setup where the model of the supervisor is unknown to the adversary but the adversary has recorded a (prefix-closed) finite set of observations of the runs of the closed-loop system. The synthesized attacker needs to ensure both the damage-reachability and the covertness against all the supervisors, which are consistent with the given set of observations. There is a gap between the de facto supremality, assuming the model of the supervisor is known, and the supremality that can be attained with a limited knowledge of the model of the supervisor, from the adversary's point of view. We consider the setup where the attacker can exercise sensor replacement/deletion attacks and actuator enablement/disablement attacks. The solution methodology proposed in this article is to reduce the synthesis of the supremal covert damage-reachable attacker, given the model of the plant and the finite set of observations, to the synthesis of the supremal safe supervisor for certain transformed plant, which shows the decidability of the observation-assisted covert attacker synthesis problem. The effectiveness of our approach is illustrated on a water tank example adapted from the literature.

The problem of covert sensor-actuator attack synthesis in the context of network discrete-event system (DES) is studied, in which the observation channel and control channel have bounded delays. We investigate a kind of attack that can 1) keep hidden, 2) implement sensor insertion, deletion and replacement attacks, and actuator enablement and disablement attacks. We propose a transformation-based method to convert the covert attack synthesis problem of network DES into the standard Ramadge-Wonham supervisory control problem. Our method is applicable to the damage-reachable goal and damage-nonblocking goal.

In this work, we address the problem of synthesis of covert attackers in the setup where the model of the plant is available, but the model of the supervisor is unknown, to the adversary. To compensate the lack of knowledge on the supervisor, we assume that the adversary has recorded a (prefix-closed) finite set of observations of the runs of the closed-loop system, which can be used for assisting the synthesis. We present a heuristic algorithm for the synthesis of covert damage-reachable attackers, based on the model of the plant and the (finite) set of observations, by a transformation into solving an instance of the partial-observation supervisor synthesis problem. The heuristic algorithm developed in this paper may allow the adversary to synthesize covert attackers without having to know the model of the supervisor, which could be hard to obtain in practice. For simplicity, we shall only consider covert attackers that are able to carry out sensor replacement attacks and actuator disablement attacks. The effectiveness of our approach is illustrated on a water tank example adapted from the literature.

Multi-agent path planning (MAPP) is crucial for large-scale mobile robot systems to work safely and properly in complex environments. Existing learning-based decentralized MAPP approaches allow each agent to gather information from nearby agents, leading to more efficient coordination among agents. However, these approaches often struggle with reasonably handling local information inputs for each agent, and their communication mechanisms between agents need to be further refined to treat those congested traffic scenarios effectively. To address these issues, we propose a decentralized MAPP approach based on imitation learning and selective communication. Our approach adopts an imitation learning architecture that enables agents to rapidly learn complex behaviors from expert planning experience. The information extraction layer is integrated with convolutional neural network (CNN) and gated recurrent unit (GRU) for capturing features from local field-of-view observations. A two-stage selective communication process based on graph attention neural network (GAT) is developed to reduce the required neighbor agents in inter-agent communication. In addition, an adaptive strategy switching mechanism utilizing local expert-planned paths is designed to support robots to escape from local traps. The effectiveness of our proposed approach is evaluated in simulated grid environments with varying map sizes, obstacle densities, and numbers of agents. Experimental results show that our approach outperforms other decentralized path planning methods in success rate while maintaining the lowest flowtime variation and communication frequency. Furthermore, our approach is computationally efficient and scalable, making it suitable for real-world applications.

Since the timed supervisory control framework was proposed, it has been adopted in many subsequent developments in various areas. Yet, the mechanism of explicitly enumerating time in terms of sequences of ticks significantly increases the number of states, causing major computational challenges in supervisor synthesis. To overcome this challenge, in this work we develop a theory of supervisory control of time-interval discrete event systems, whereby intervals are used to compactly encode sequences of ticks. By introducing the concepts of freeand coercive(interval) languages, we are able to describe the impact of event forcing (i.e., time preemption) in a resulting time-interval language, upon which we introduce controllability property. It turns out that coercive languages are not closed under union, in contrast to free languages. The supremal controllable coercive sublanguage for a given time-interval plant and time-interval requirement exists, which can be computed by a synthesis algorithm presented in this paper.

In this work, we investigate the problem of covert attacker synthesis for supervisory control of discrete-event systems. In particular, we consider attackers that exercise both actuator attacks and sensor attacks, where the (partial-observation) attackers may or may not eavesdrop the control commands issued by the supervisor. We develop a reduction from the covert attacker synthesis problem to the Ramadge-Wonham supervisor synthesis problem, which generalizes our previous work on a reduction based approach for covert actuator attacker synthesis.

A reduction of a source distribution is a collection of smaller sized distributions that are collectively equivalent to the source distribution with respect to the property of decomposability. That is, an arbitrary language is decomposable with respect to the source distribution if and only if it is decomposable with respect to each smaller sized distribution (in the reduction). The notion of reduction of distributions has previously been proposed to improve the complexity of decomposability verification. In this paper, we address the problem of generating (optimal) reductions of distributions automatically. A (partial) solution to this problem is provided, which consists of an incremental algorithm for the production of candidate reductions and a reduction validation procedure. In the incremental production stage, backtracking is applied whenever a candidate reduction that cannot be validated is produced. A strengthened substitution-based proof technique is used for reduction validation, while a fixed template of candidate counter examples is used for reduction refutation; put together, they constitute our (partial) solution to the reduction verification problem. In addition, we show that a recursive approach for the generation of (small) reductions is easily supported. [ABSTRACT FROM AUTHOR]

With a growing number of interacting modules in today's complex systems, finding an effective tool for detecting design errors has become increasingly challenging. We present procedures for analyzing blocking or deadlock in systems consisting of isomorphic modules instantiated from a template. These procedures identify all possible blocking or deadlock sources and directly compute the maximum number of isomorphic modules under which the system is guaranteed to be nonblocking or deadlock-free. [ABSTRACT FROM AUTHOR]

In this paper, we outline a method for carrying out efficient (max, +) matrix multiplication when using the heaps of pieces framework. We present an algorithm for multiplying an arbitrary mby rmatrix Xby a rby rheaps of pieces matrix M, making it possible to calculate the resulting matrix in worst case time complexity O(mr),rather than O(mr2)which is required when using the matrix multiplication definition. We also give an algorithm for multiplying Mby an arbitrary rby nmatrix Xwith worst case time complexity O(nr).Finally, we consider a variant of the standard heaps of pieces model, and present an improved matrix multiplication algorithm for this variant as well.

In this work, a notion of reduction of distributions is proposed as a technical tool for improving the complexity of decomposability verification and supporting parallel verification of decomposability, by exploiting the rich structures of distributions. We provide some results that reduce the search space of candidate reductions, as a first step toward efficiently computing optimal reductions. It is then shown that a distribution has a reduction if and only if a particular candidate reduction is indeed a reduction. We then provide a sound substitution-based proof technique that can be used for (automatic) reduction verification. Techniques for refuting candidate reductions are also provided. We then explain an application of the decomposability verification problem in the lower bound proofs for the problem of supervisor decomposition and the problem of existence of a decentralized supervisor. Finally, some other applications of the notion of reduction of distributions are also shown. [ABSTRACT FROM PUBLISHER]

It is shown that the problem whether an arbitrary regular language has a non-empty decomposable sublanguage with respect to a fixed distribution is decidable if and only if the independence relation induced by the distribution is transitive. A sufficient condition on the distributed control architecture is then derived, under which there exist some fixed non-blocking local generators such that the distributed supervisor synthesis problem is undecidable. We also show that a natural formulation of the parameterized supervisor synthesis problem is undecidable for a fixed non-blocking generator template, so long as the template alphabet has at least two private events and one global event that are controllable. In particular, all the undecidability results are still valid even if star free specification languages are considered. [ABSTRACT FROM AUTHOR]

Astronauts could be exposed to solar particle event (SPE) radiation, which is comprised mostly of proton radiation. Proton radiation is also a treatment option for certain cancers. Both astronauts and clinical patients exposed to ionizing radiation are at risk for loss of white blood cells (WBCs), which are the body's main defense against infection. In this report, the effect of Neulasta treatment, a granulocyte colony stimulating factor, after proton radiation exposure is discussed. Mini pigs exposed to total body proton irradiation at a dose of 2 Gy received 4 treatments of either Neulasta or saline injections. Peripheral blood cell counts and thromboelastography parameters were recorded up to 30 days post-irradiation. Neulasta significantly improved WBC loss, specifically neutrophils, in irradiated animals by approximately 60% three days after the first injection, compared to the saline treated, irradiated animals. Blood cell counts quickly decreased after the last Neulasta injection, suggesting a transient effect on WBC stimulation. Statistically significant changes in hemostasis parameters were observed after proton radiation exposure in both the saline and Neulasta treated irradiated groups, as well as internal organ complications such as pulmonary changes. In conclusion, Neulasta treatment temporarily alleviates proton radiation-induced WBC loss, but has no effect on altered hemostatic responses.

During a major solar particle event (SPE), astronauts in space are at risk of exposure to an increased dose of proton radiation. The whole body distribution of the absorbed SPE proton dose is inhomogeneous, and such an inhomogeneous SPE proton dose can be simulated by electron radiation. Using Yucatan minipigs as an animal model, we compared the time courses of leukocyte count changes after exposure to proton simulated SPE (pSPE) radiation or electron simulated SPE (eSPE) radiation. The results demonstrated that the time required after irradiation to reach the lowest leukocyte counts was generally comparable between the pSPE and eSPE radiation exposures. However, the leukocyte count often recovered faster after electron irradiation compared to proton irradiation at the corresponding doses. In addition, the radiation dose required to achieve comparable magnitudes of leukocyte count decrease was higher in the eSPE animals than for the pSPE animals. In conclusion, based on the magnitude of the decrease and the time required to reach the lowest leukocyte counts after irradiation, the pSPE radiation was more effective than the eSPE radiation in reducing the peripheral leukocyte counts. Lymphocytes appeared to be the most sensitive type of leukocytes in response to either type of SPE radiation. It is particularly noteworthy that following exposure to pSPE radiation at the skin doses >5 Gy, the neutrophils do not recover from the radiation damage at times up to 30 days, and the neutrophils have not recovered to their baseline levels even at 90 days post-irradiation. These results suggest a marked difference in the ability of the neutrophils to recover from pSPE radiation compared with the results observed for eSPE radiation.

This paper considers integrated target search, tasking and tracking using multiple fixed-wing UAVs in urban environments. The problem is to design autonomy for each individual UAV autonomous and distributed tasking. Control logic design based on finite state automaton (FSA) model, integrating the four modes of operations, i.e., takeoff mode, fly-to-AO (area of operation) mode, search mode and tracking mode, is developed. An efficient distributed multi-UAV target search algorithm is also presented. UAV guidance and control is built based on combined urban road map and target detection probability map information. For target tracking, by using geometric relations (relative position, orientations, speed ratio, and minimal turning radius), a systematic algorithm is developed to generate an optimal path online for a fixed-wing UAV to track a moving target. In addition, control method for a group of UAVs to keep track a target convoy is also addressed. Finally, the proposed decentralized algorithms are evaluated by simulations adopting a real UAV model.

Immune system adaptation during spaceflight is a concern in space medicine. Decreased circulating leukocytes observed during and after space flight infer suppressed immune responses and susceptibility to infection. The microgravity aspect of the space environment has been simulated on Earth to study adverse biological effects in astronauts. In this report, the hindlimb unloading (HU) model was employed to investigate the combined effects of solar particle event-like proton radiation and simulated microgravity on immune cell parameters, including lymphocyte subtype populations and activity. Lymphocytes are a type of white blood cell critical for adaptive immune responses and T lymphocytes are regulators of cell-mediated immunity, controlling the entire immune response. Mice were suspended prior to and after proton radiation exposure (0 or 2 Gy doses) and total leukocyte numbers and splenic lymphocyte functionality were evaluated on days 4 or 21 after combined HU and radiation exposure. Total white blood cell (WBC), lymphocyte, neutrophil and monocyte counts are reduced by ∼65, 70, 55 and 70%, respectively, compared with the non-treated control group 4 days after combined exposure. Splenic lymphocyte subpopulations are altered at both time points investigated. At 21 days post-exposure to combined HU and proton radiation, T-cell activation and proliferation were assessed in isolated lymphocytes. Cell surface expression of the Early Activation Marker, CD69, is decreased by 30% in the combined treatment group, compared with the non-treated control group and cell proliferation was suppressed by ∼50%, compared with the non-treated control group. These findings reveal that the combined stressors (HU and proton radiation exposure) induce decreased leukocyte numbers and function, contributing to immune system dysfunction in crew members. This research was supported by the NIH Training Grant 2T32CAO9677 and the National Space Biomedical Research Institute (NSBRI) Center of Acute Radiation Research (CARR) grant. The NSBRI is funded through the National Aeronautics and Space Administration (NASA) Class Code (NCC) 9-58. The authors declare that this work has been published: Sanzari JK, Romero-Weaver AL, James G, Krigsfeld G, Lin L, Diffenderfer ES, Kennedy AR. Leukocyte activity is altered in a ground based murine model of microgravity and proton radiation exposure. PLoS One. 2013;8(8):e71757. doi:10.1371/journal.pone.0071757.