Recent studies in sequence-to-sequence learning demonstrate that RNN encoder-decoder structure can successfully generate Chinese poetry. However, existing methods can only generate poetry with a given first line or user's intent theme. In this paper, we proposed a three-stage multi-modal Chinese poetry generation approach. Given a picture, the first line, the title and the other lines of the poem are successively generated in three stages. According to the characteristics of Chinese poems, we propose a hierarchy-attention seq2seq model which can effectively capture character, phrase, and sentence information between contexts and improve the symmetry delivered in poems. In addition, the Latent Dirichlet allocation (LDA) model is utilized for title generation and improve the relevance of the whole poem and the title. Compared with strong baseline, the experimental results demonstrate the effectiveness of our approach, using machine evaluations as well as human judgments.
This paper studies an extension and improvement of the joint detection-decoding algorithm for nonbinary LDPC-coded modulation systems. The iterative joint detection-decoding (IJDD) algorithm in [1] combines nonbinary LDPC decoding with signal detection based on the hard-message passing strategy, resulting in significantly reduced decoding complexity. However, it applies only to majority-logic decodable nonbinary LDPC codes with high column weight. For nonbinary LDPC codes with low column weight, a noticeable performance loss will be incurred. To handle this problem, we propose a reliability-based iterative joint detection-decoding (also termed improved IJDD) algorithm, which combines the accumulated reliability of symbols based on the one-step majority-logic decoding (MLGD) algorithm and a Chase-like local list decoding algorithm. Simulation results show that the improved IJDD algorithm outperforms the IJDD algorithm by about 0.3 dB using nonbinary LDPC codes with high column weight, and by about 3 dB using nonbinary LDPC codes with low column weight (d v = 4), while maintaining the low complexity of decoding. Compared to the FFT-QSPA, the proposed algorithm has a performance degradation of 0.5 dB in the high column weight regime, and about 1 dB in the low column weight regime.
We present a microfluidic technique for measuring the deformability of single cells using the pressure required to deform such cells through micrometre-scale tapered constrictions. Our technique is equivalent to whole-cell micropipette aspiration, but involves considerably simpler operation, less specialized equipment, and less technical skill. Single cells are infused into a microfluidic channel, and then deformed through a series of funnel-shaped constrictions. The constriction openings are sized to create a temporary seal with each cell as it passes through the constriction, replicating the interaction with the orifice of a micropipette. Precisely controlled deformation pressures are generated using an external source and then attenuated 100 : 1 using an on-chip microfluidic circuit. Our apparatus is capable of generating precisely controlled pressures as small as 0.3 Pa in a closed microchannel network, which is impervious to evaporative losses that normally limit the precision of such equipment. Intrinsic cell deformability, expressed as cortical tension, is determined from the threshold deformation pressure using the liquid-drop model. We measured the deformability of several types of nucleated cells and determined the optimal range of constriction openings. The cortical tension of passive human neutrophils was measured to be 37.0 ± 4.8 pN μm−1, which is consistent with previous micropipette aspiration studies. The cortical tensions of human lymphocytes, RT4 human bladder cancer cells, and L1210 mouse lymphoma cells were measured to be 74.7 ± 9.8, 185.4 ± 25.3, and 235.4 ± 31.0 pN μm−1 respectively. The precision and usability of our technique demonstrates its potential as a biomechanical assay for wide-spread use in biological and clinical laboratories.
Nuclear medicine imaging plays a pivotal role in the diagnosis of systemic bone diseases. However, conventional discrete models struggle to handle the inherent fuzziness and randomness of such images. Recently proposed, the neural memory ordinary differential equation (nmODE) model exhibits intriguing characteristics and has shown significant improvements in classification tasks. In this study, we leverage the powerful nonlinear and dynamic modeling capabilities of nmODE and extend its application to the field of nuclear medicine image segmentation. The experimental results affirm that the inherent stability of nmODE is highly suitable for addressing the task of lesion segmentation in nuclear medicine images characterized by indistinctness and stochasticity.
Circulating tumor cells (CTCs) offer tremendous potential for the detection and characterization of cancer. A key challenge for their isolation and subsequent analysis is the extreme rarity of these cells in circulation. Here, a novel label‐free method is described to enrich viable CTCs directly from whole blood based on their distinct deformability relative to hematological cells. This mechanism leverages the deformation of single cells through tapered micrometer scale constrictions using oscillatory flow in order to generate a ratcheting effect that produces distinct flow paths for CTCs, leukocytes, and erythrocytes. A label‐free separation of circulating tumor cells from whole blood is demonstrated, where target cells can be separated from background cells based on deformability despite their nearly identical size. In doping experiments, this microfluidic device is able to capture >90% of cancer cells from unprocessed whole blood to achieve 10 4 ‐fold enrichment of target cells relative to leukocytes. In patients with metastatic castration‐resistant prostate cancer, where CTCs are not significantly larger than leukocytes, CTCs can be captured based on deformability at 25× greater yield than with the conventional CellSearch system. Finally, the CTCs separated using this approach are collected in suspension and are available for downstream molecular characterization.
Ligand-modified nanoparticles (NPs) are an effective tool to increase the endocytosis efficiency of drugs, but these functionalized NPs face the drawback of "easy uptake hard transcytosis" in the oral delivery of proteins and peptides. Adversely, the resulting deficiency in transcytosis has not attracted much attention. Herein, NPs modified with the low-density lipoprotein receptor (LDLR) ligand NH2-C6-[cMPRLRGC]c-NH2, i.e., peptide-22 (P22NPs) were fabricated to investigate strategies related to the enhancement of transcytosis. By systematically studying the intracellular trafficking of NPs, it was found that reduced transcytosis might be associated with the entrapment of P22NPs in endosomes or lysosomes and limited basolateral exocytosis. On this basis, the prevention of the endolysosomal entrapment of NPs and the acceleration of basolateral exocytosis should be considered as strategies to enhance the transcytosis of NPs. By screening chemicals that could help the endosomal/lysosomal escape of chemicals related to LDLR-mediated transcytosis, it was shown that hemagglutinin-2 (HA2) and metformin had higher abilities to enhance the exocytosis of P22NPs. The transcytosis efficiencies of insulin loaded in P22NPs were also investigated, and a 3.2-fold increase in transcytosis was observed in comparison with free insulin. The transcytosis efficiencies of insulin could be further increased by the addition of metformin or HA2 (3.6-fold or 4.1-fold higher than that of free insulin). Inspiringly, the simultaneous addition of the abovementioned two chemicals led to the highest transcytosis efficiency of insulin, which was up to 5.1-fold higher than that of free insulin. These results demonstrated that endolysosomal entrapment and basolateral exocytosis are two of the most important limiting steps for the "easy uptake hard transcytosis" of orally administered ligand-modified NPs. Moreover, our work provides a new point of view for the design of novel oral drug delivery systems.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
