Search

Background: The aim of this study was to measure serum brain injury biomarkers in patients with COVID-19 admitted to intensive care unit (ICU), without evidence of brain impairment, and to determine potential correlations with systemic inflammatory markers, illness severity, and outcome. Methods: In patients admitted to the ICU with COVID-19, without clinical evidence of brain injury, blood S100 calcium-binding protein B (S100B), neuron-specific enolase (NSE) and interleukin-6 (IL-6) were measured on admission. Clinical, routine laboratory data and illness severity were recorded. Comparisons between 28-day survivors and non-survivors and correlations of neurological biomarkers to other laboratory data and illness severity, were analyzed. Results: We included 50 patients, median age 64 [IQR 58–78] years, 39 (78%) males, 39 (78%) mechanically ventilated and 11 (22%) under high flow nasal oxygen treatment. S100B and NSE were increased in 19 (38%) and 45 (90%) patients, respectively. S100B was significantly elevated in non-survivors compared to survivors: 0.15 [0.10–0.29] versus 0.11 [0.07–0.17] μg/L, respectively, (p = 0.03), and significantly correlated with age, IL-6, arterial lactate, noradrenaline dose, illness severity and lymphocyte count. IL-6 was significantly correlated with C-reactive protein, noradrenaline dose and organ failure severity. NSE was correlated only with lactate dehydrogenase. Conclusion: Brain injury biomarkers were frequently elevated in COVID-19 ICU patients, in the absence of clinical evidence of brain injury. S100B was significantly correlated with IL-6, low lymphocyte count, hypoperfusion indices, illness severity, and short-term outcome. These findings indicate a possible brain astrocytes and neurons involvement, also suggesting a broader role of S100B in systemic inflammatory response.

The Internet of Battlefield Things is a newly born cyberphysical system and, even though it shares a lot with the Internet of Things and with ad hoc networking, substantial research is required to cope with its scale and peculiarities. This article examines a fundamental problem pertaining to the routing of information, i.e., the calculation of a backbone network. We model an IoBT network as a network with multiple layers and employ the concept of domination for multilayer networks. This is a significant departure from earlier works, and in spite of the huge literature on the topic during the past twenty years, the problem in IoBT networks is different since these networks are multilayer networks, thus making inappropriate all the past, related literature because it deals with single layer (flat) networks. We establish the computational complexity of our problem, and design a distributed algorithm for computing connected dominating sets with small cardinality. We analyze the performance of the proposed algorithm on generated topologies, and compare it against two—the only existing—competitors. The proposed algorithm establishes itself as the clear winner in all experiments concerning the dominating set from a size-wise and an energy-wise perspective achieving a performance gain of about 15%.

Packet scheduling/routing in wireless ad hoc networks is a fundamental problem for ad hoc networking. Backpressure routing is a solid and throughput optimal policy for such networks, but suffers from increased delays. In this article, we present two holistic approaches to improve upon the delay problems of backpressuretype algorithms. We develop two scheduling algorithms, namely Voting backpressure and Layered backpressure routing, which are throughput optimal. We experimentally compare the proposed algorithms against state-of-the-art delay-aware backpressure algorithms, which provide optimal throughput, for different payloads and network topologies, both for static and mobile networks. The experimental evaluation of the proposed delay reduction algorithms attest their superiority in terms of QoS, robustness, low computational complexity and simplicity.