Search for: All records

Pick’s disease (PiD) is a subtype of the tauopathy form of frontotemporal lobar degeneration (FTLD-tau) characterized by intraneuronal 3R-tau inclusions. PiD can underly various dementia syndromes, including primary progressive aphasia (PPA), characterized by an isolated and progressive impairment of language and left-predominant atrophy, and behavioral variant frontotemporal dementia (bvFTD), characterized by progressive dysfunction in personality and bilateral frontotemporal atrophy. In this study, we investigated the neocortical and hippocampal distributions of Pick bodies in bvFTD and PPA to establish clinicopathologic concordance between PiD and the salience of the aphasicversusbehavioral phenotype. Eighteen right-handed cases with PiD as the primary pathologic diagnosis were identified from the Northwestern University Alzheimer’s Disease Research Center brain bank (bvFTD, N = 9; PPA, N = 9). Paraffin-embedded sections were stained immunohistochemically with AT8 to visualize Pick bodies, and unbiased stereological analysis was performed in up to six regions bilaterally [middle frontal gyrus (MFG), superior temporal gyrus (STG), inferior parietal lobule (IPL), anterior temporal lobe (ATL), dentate gyrus (DG) and CA1 of the hippocampus], and unilateral occipital cortex (OCC). In bvFTD, peak neocortical densities of Pick bodies were in the MFG, while the ATL was the most affected in PPA. Both the IPL and STG had greater leftward pathology in PPA, with the latter reaching significance (p < 0.01). In bvFTD, Pick body densities were significantly right-asymmetric in the STG (p < 0.05). Hippocampal burden was not clinicopathologically concordant, as both bvFTD and PPA cases demonstrated significant hippocampal pathology compared to neocortical densities (p < 0.0001). Inclusion-to-neuron analyses in a subset of PPA cases confirmed that neurons in the DG are disproportionately burdened with inclusions compared to neocortical areas. Overall, stereological quantitation suggests that the distribution of neocortical Pick body pathology is concordant with salient clinical features unique to PPA vs. bvFTD while raising intriguing questions about the selective vulnerability of the hippocampus to 3R-tauopathies.

 

In the advancing field of 5G technologies, particularly at the 60 GHz band, dielectric resonator antennas (DRAs) stand out for their low conduction loss and high radiation efficiency. However, the traditional design process for DRAs, predominantly reliant on intuitive reasoning and trial‐and‐error methods, is notably inefficient and resource‐intensive. Addressing this critical challenge, our research introduces a pioneering approach: a generative adversarial network (GAN)‐based model specifically tailored for automating DRA structure design. This novel model represents the first of its kind in the domain, marking a significant departure from conventional methods. Our GAN model uniquely integrates a simulator for DRA modeling and a generator for DRA structure design, streamlining the design process. To effectively train this model, we created a simulated data set comprising pattern–annotation pairs of geometric shapes andS₁₁parameters. This data set enabled the GAN to capture the intrinsic principles underlying DRA design. The practical impact of our model is profound; it significantly expedites the DRA design process, aligning it more closely with specific user requirements while conserving valuable time and resources. This breakthrough approach not only enhances the efficiency of DRA design but also sets a new standard in antenna technology development for future wireless communications.

 

As the size of data generated every day grows dramatically, the computational bottleneck of computer systems has shifted toward storage devices. The interface between the storage and the computational platforms has become the main limitation due to its limited bandwidth, which does not scale when the number of storage devices increases. Interconnect networks do not provide simultaneous access to all storage devices and thus limit the performance of the system when executing independent operations on different storage devices. Offloading the computations to the storage devices eliminates the burden of data transfer from the interconnects. Near-storage computing offloads a portion of computations to the storage devices to accelerate big data applications. In this article, we propose a generic near-storage sort accelerator for data analytics, NASCENT2, which utilizes Samsung SmartSSD, an NVMe flash drive with an on-board FPGA chip that processes data in situ. NASCENT2 consists of dictionary decoder, sort, and shuffle FPGA-based accelerators to support sorting database tables based on a key column with any arbitrary data type. It exploits data partitioning applied by data processing management systems, such as SparkSQL, to breakdown the sort operations on colossal tables to multiple sort operations on smaller tables. NASCENT2 generic sort provides 2 × speedup and 15.2 × energy efficiency improvement as compared to the CPU baseline. It moreover considers the specifications of the SmartSSD (e.g., the FPGA resources, interconnect network, and solid-state drive bandwidth) to increase the scalability of computer systems as the number of storage devices increases. With 12 SmartSSDs, NASCENT2 is 9.9× (137.2 ×) faster and 7.3 × (119.2 ×) more energy efficient in sorting the largest tables of TPCC and TPCH benchmarks than the FPGA (CPU) baseline.