Edge computing provides off-load computing and application services close to end-users, greatly reducing cloud pressure and communication overhead. However, wireless edge networks still face the risk of network attacks. To ensure the security of wireless edge networks, we present Federated Learning-based Attention Gated Recurrent Unit (FedAGRU), an intrusion detection algorithm for wireless edge networks. FedAGRU differs from current centralized learning methods by updating universal learning models rather than directly sharing raw data among edge devices and a central server. We also apply the attention mechanism to increase the weight of important devices, by avoiding the upload of unimportant updates to the server, FedAGRU can greatly reduce communication overhead while ensuring learning convergence. Our experimental results show that, compared with other centralized learning algorithms, FedAGRU improves detection accuracy by approximately 8%. In addition, FedAGRU's communication cost is 70% less than other federated learning algorithms, and it exhibits strong robustness against poisoning attacks. INDEX TERMS wireless edge, intrusion detection, federated learning, gated recurrent unit

BackgroundGrowth of nerve fibers has been shown to occur in a rabbit model of intravertebral disc degeneration (IVD) induced by needle puncture. As nerve growth may underlie the process of chronic pain in humans affected by disc degeneration, we sought to investigate the factors underlying nerve ingrowth in a minimally invasive annulotomy rabbit model of IVD by comparing the effects of empty disc defects with those of defects filled with poly(lactic-co-glycolic acid)/fibrin gel (PLGA) plugs.MethodsNew Zealand white rabbits (n = 24) received annular injuries at three lumbar levels (L3/4, L4/5, and L5/6). The discs were randomly assigned to four groups: (a) annular defect (1.8-mm diameter; 4-mm depth) by mini-trephine, (b) annular defect implanted with a PLGA scaffold containing a fibrin gel, (c) annular puncture by a 16G needle (5-mm depth), and (d) uninjured L2/3 disc (control). Disc degeneration was evaluated by radiography, MRI, histology, real-time PCR, and analysis of proteoglycan (PG) content. Nerve ingrowth into the discs was assessed by immunostaining with the nerve marker protein gene product 9.5.ResultsInjured discs showed a progressive disc space narrowing with significant disc degeneration and proteoglycan loss, as confirmed by imaging results, molecular and compositional analysis, and histological examinations. In 16G punctured discs, nerve ingrowth was observed on the surface of scar tissue. In annular defects, nerve fibers were found to be distributed along small fissures within the fibrocartilaginous-like tissue that filled the AF. In discs filled with PLGA/ fibrin gel, more nerve fibers were observed growing deeper into the inner AF along the open annular track.  In addition, innervations scores showed significantly higher than those of punctured discs and empty defects. A limited vascular proliferation was found in the injured sites and regenerated tissues.ConclusionsNerve ingrowth was significantly higher in PLGA/fibrin-filled discs than in empty defects. Possible explanations include (i) annular fissures along the defect and early loss of proteoglycan may facilitate the ingrowth process and (ii) biodegradable PLGA/fibrin gel may promote adverse growth of nerves and blood vessels into deeper parts of injured disc. The rabbit annular defect model of disc degeneration appears suitable to investigate the effects of nerve ingrowth in relation to pain generation.

No abstract

No abstract

Objective: We aimed to explore the associated clinical phenotype and the natural history of patients with SYNGAP1 gene variations during early childhood and to identify their genotype–phenotype correlations.Methods: This study used a cohort of 13 patients with epilepsy and developmental disorder due to SYNGAP1 mutations, namely, 7 patients from Shenzhen Children’s Hospital between September 2014 and January 2020 and 6 patients from previously published studies. Their clinical data were studied.Results: A total of 13 children with SYNGAP1 gene variants (eight boys and five girls) were identified. The age of disease onset was in infancy. Mutations were located between exons 8 and 15; most were frameshift or truncated mutations. Four mutation sites (c.924G > A, c.1532-2_1532del, c.1747_1755dup, and c.1735_1738del) had not been reported before. All patients had global developmental delay within the first year of life, and intellectual impairment became gradually apparent. Some of them developed behavioral problems. The developmental delay occurred before the onset of seizures. All seven patients in our cohort presented with epilepsy; myoclonic seizures, absence seizures, and epileptic spasms were the most common seizure types. Abnormal electroencephalograms were identified from five patients before the onset of their seizures. All patients suffered from drug-resistance seizures. However, comorbidities such as behavioral problems were less frequently observed.Conclusion: The most common age of disease onset in SYNGAP1 gene mutations is in infancy, while neurodevelopmental delay and epilepsy are the major phenotypes. They have a higher percentage of drug-resistant epilepsy and epileptic spasms than those in previous reports. We should give attention to the patients with abnormal EEGs without seizures and think about the suitable time of the anti-seizure medications for them. We have not found the genotype–phenotype correlation.Trial registration: Chinese Clinical Trial Registry, Registration number: ChiCTR2100049289 (https://www.chictr.org.cn/listbycreater.aspx).