Search

Abstract The complexity of neuronal networks, characterized by interconnected neurons, presents significant challenges in control due to their nonlinear and intricate behaviour. This paper introduces a novel method designed to generate control inputs for neuronal networks to regulate the firing patterns of modules within the network. This methodology is built upon temporal deep unfolding‐based model predictive control, a technique rooted in the deep unfolding method commonly used in wireless signal processing. To address the unique dynamics of neurons, such as zero gradients in firing times, the method employs approximations of input currents using a sigmoid function during its development. The effectiveness of this approach is validated through extensive numerical simulations. Furthermore, control experiments were conducted by reducing the number of input neurons to identify critical features for control. Various selection techniques were utilized to pinpoint key input neurons. These experiments shed light on the importance of specific input neurons in controlling module firing within neuronal networks. Thus, this study presents a tailored methodology for managing networked neurons, extends temporal deep unfolding‐based model predictive control to nonlinear systems with reset dynamics, and demonstrates its ability to achieve desired firing patterns in neuronal networks.

In the context of the theory of multi-agent systems, the shepherding problem refers to designing the dynamics of a herding agent, called a sheepdog, so that a given flock of agents, called sheep, is guided into a goal region. Although several effective methodologies and algorithms have been proposed in the last decade for the shepherding problem under various formulations, little research has been directed to the practically important case in which the flock contains sheep agents unresponsive to the sheepdog agent. To fill in this gap, we propose a sheepdog algorithm for guiding unresponsive sheep in this paper. In the algorithm, the sheepdog iteratively applies an existing shepherding algorithm, the farthest-agent targeting algorithm, while dynamically switching its destination. This procedure achieves the incremental growth of a controllable flock, which finally enables the sheepdog to guide the entire flock into the goal region. Furthermore, we illustrate by numerical simulations that the proposed algorithm can outperform the farthest-agent targeting algorithm.

Abstract This paper proposes a computational technique to solve discrete‐time optimal control problems for nonlinear systems subject to stochastic disturbances, hard input constraints, and soft constraints. The approach employs the idea of deep unfolding, which is a recently developed model‐based deep learning method that is applicable to iterative algorithms. By regarding each state transition of the dynamical system as an iteration step, these state transitions are unfolded into the layers of a deep neural network. This produces a computational graph that contains trainable parameters that determine control inputs. These parameters are optimized by training the deep neural network using the standard deep learning technique of backpropagation and incremental training. Then, the optimal control inputs are obtained from the optimized parameters. This provides a way to optimize the control inputs that otherwise would be difficult to obtain. The effectiveness of the proposed technique is demonstrated by numerical experiments for the maximum hands‐off control problem and constrained model predictive control problems using a highly nonlinear model of a continuous stirred tank reactor.

In this paper, we study the problem of optimizing the stability of positive semi-Markov jump linear systems. We specifically consider the problems of tuning the coefficients of the system matrices for maximizing the exponential decay rate of the system under a budget-constraint and minimizing the parameter tuning cost under the decay rate constraint. By using a result from the matrix theory on the log-log convexity of the spectral radius of nonnegative matrices, we show that the stability optimization problems are reduced to convex optimization problems under certain regularity conditions on the system matrices and the cost function. We illustrate the validity and effectiveness of the proposed results by using an example from the population biology.

In this study, we analyzed the problem of accelerating the linear average consensus algorithm for complex networks. We propose a data-driven approach to tuning the weights of temporal (i.e., time-varying) networks using deep learning techniques. Given a finite-time window, the proposed approach first unfolds the linear average consensus protocol to obtain a feedforward signal-flow graph, which is regarded as a neural network. The edge weights of the obtained neural network are then trained using standard deep learning techniques to minimize consensus error over a given finite-time window. Through this training process, we obtain a set of optimized time-varying weights, which yield faster consensus for a complex network. We also demonstrate that the proposed approach can be extended for infinite-time window problems. Numerical experiments revealed that our approach can achieve a significantly smaller consensus error compared to baseline strategies.

BackgroundPneumonia is a common cause of illness and death of the elderly in Japan. Its prevalence is escalating globally with the aging of population. To describe the latest trends in pneumonia hospitalizations, especially aspiration pneumonia (AP) cases, we assessed the clinical records of pneumonia patients admitted to core acute care hospitals in Miyagi prefecture, Japan.MethodsA retrospective multi-institutional joint research was conducted for hospitalized pneumonia patients aged ≥20 years from January 2019 to December 2019. Clinical data of patients were collected from the medical records of eight acute care hospitals.ResultsOut of the 1,800 patients included in this study, 79% of the hospitalized pneumonia patients were aged above 70 years. The most common age group was in the 80s. The ratio of AP to total pneumonia cases increased with age, and 692 out of 1,800 patients had AP. In univariate analysis, these patients had significantly older ages, lower body mass index (BMI), a lower ratio of normal diet intake and homestay before hospitalization, along with more AP recurrences and comorbidities. During hospitalization, AP patients had extended fasting periods, more swallowing assessments and interventions, longer hospitalization, and higher in-hospital mortality rate than non-AP patients. A total of 7% and 2% AP patients underwent video endoscopy and video fluorography respectively. In multivariate analysis, lower BMI, lower C-reactive protein, a lower ratio of homestay before hospitalization, a higher complication rate of cerebrovascular disease, dementia, and neuromuscular disease were noted as a characteristic of AP patients. Swallowing interventions were performed for 51% of the AP patients who had been hospitalized for more than two weeks. In univariate analysis, swallowing intervention improved in-hospital mortality. Lower AP recurrence before hospitalization and a lower ratio of homestay before hospitalization were indicated as characteristics of AP patients of the swallowing intervention group from multivariate analysis. Change in dietary pattern from normal to modified diet was observed more frequently in the swallowing intervention group.ConclusionAP accounts for 38.4% of all pneumonia cases in acute care hospitals in Northern Japan. The use of swallowing evaluations and interventions, which may reduce the risk of dysphagia and may associate with lowering mortality in AP patients, is still not widespread.

A fairly quite rare case of osseous tumor arising from the Eustachian tube (ET) is described. A 56-year-old man presented with a smooth bulky mass in the nasopharynx and secretory otitis media in the right ear. Computed tomography and magnetic resonance imaging indicated a solid tumor-like region occupying the nasopharynx with apparent extension to the right ET. Transnasal endoscopic surgery demonstrated that the tumor had originated from the ET, and the tumor was partially removed in the area where it had expanded into the nasopharynx. The pathological diagnosis was an osseous hamartoma consisting of bony tissue, fat cells and fibroblasts. To our knowledge, this is the first reported case of osseous hamartoma arising from the ET in the world literature. The patient presented symptoms of nasal obstruction and unilateral aural fullness. Reduction surgery and tympanostomy tube insertion were useful for accurate diagnosis and resolution of the symptoms.