Real-world prediction of preclinical Alzheimer's disease with a deep generative model.

Amyloid positivity is an early indicator of Alzheimer's disease and is necessary to determine the disease. In this study, a deep generative model is utilized to predict the amyloid positivity of cognitively normal individuals using proxy measures, such as structural MRI scans, demographic variables, and cognitive scores, instead of invasive direct measurements. Through its remarkable efficacy in handling imperfect datasets caused by missing data or labels, and imbalanced classes, the model outperforms previous studies and widely used machine learning approaches with an AUROC of 0.8609. Furthermore, this study illuminates the model's adaptability to diverse clinical scenarios, even when feature sets or diagnostic criteria differ from the training data. We identify the brain regions and variables that contribute most to classification, including the lateral occipital lobes, posterior temporal lobe, and APOE ϵ4 allele. Taking advantage of deep generative models, our approach can not only provide inexpensive, non-invasive, and accurate diagnostics for preclinical Alzheimer's disease, but also meet real-world requirements for clinical translation of a deep learning model, including transferability and interpretability.

Convolutional Hough Matching Networks for Robust and Efficient Visual Correspondence.

Despite advances in feature representation, leveraging geometric relations is crucial for establishing reliable visual correspondences under large variations of images. In this work we introduce a Hough transform perspective on convolutional matching and propose an effective geometric matching algorithm, dubbed Convolutional Hough Matching (CHM). The method distributes similarities of candidate matches over a geometric transformation space and evaluates them in a convolutional manner. We cast it into a trainable neural layer with a semi-isotropic high-dimensional kernel, which learns non-rigid matching with a small number of interpretable parameters. To further improve the efficiency of high-dimensional voting, we also propose to use an efficient kernel decomposition with center-pivot neighbors, which significantly sparsifies the proposed semi-isotropic kernels without performance degradation. To validate the proposed techniques, we develop the neural network with CHM layers that perform convolutional matching in the space of translation and scaling. Our method sets a new state of the art on standard benchmarks for semantic visual correspondence, proving its strong robustness to challenging intra-class variations.

Harmonization of Multicenter Cortical Thickness Data by Linear Mixed Effect Model.

Objective: Analyzing neuroimages being useful method in the field of neuroscience and neurology and solving the incompatibilities across protocols and vendors have become a major problem. We referred to this incompatibility as "center effects," and in this study, we attempted to correct such center effects of cortical feature obtained from multicenter magnetic resonance images (MRIs). Methods: For MRI of a total of 4,321 multicenter subjects, the harmonized w-score was calculated by correcting biological covariates such as age, sex, years of education, and intercranial volume (ICV) as fixed effects and center information as a random effect. Afterward, we performed classification tasks using principal component analysis (PCA) and linear discriminant analysis (LDA) to check whether the center effect was successfully corrected from the harmonized w-score. Results: First, an experiment was conducted to predict the dataset origin of a random subject sampled from two different datasets, and it was confirmed that the prediction accuracy of linear mixed effect (LME) model-based w-score was significantly closer to the baseline than that of raw cortical thickness. As a second experiment, we classified the data of the normal and patient groups of each dataset, and LME model-based w-score, which is biological-feature-corrected values, showed higher classification accuracy than the raw cortical thickness data. Afterward, to verify the compatibility of the dataset used for LME model training and the dataset that is not, intraobject comparison and w-score RMSE calculation process were performed. Conclusion: Through comparison between the LME model-based w-score and existing methods and several classification tasks, we showed that the LME model-based w-score sufficiently corrects the center effects while preserving the disease effects from the dataset. We also showed that the preserved disease effects have a match with well-known disease atrophy patterns such as Alzheimer's disease or Parkinson's disease. Finally, through intrasubject comparison, we found that the difference between centers decreases in the LME model-based w-score compared with the raw cortical thickness and thus showed that our model well-harmonizes the data that are not used for the model training.

Selective Production of Bio-Based Linear Alpha-Olefin from Wasted Fatty Alcohol on Al2O3 for Bio-Based Chemicals.

The catalytic dehydration of a bio-based fatty alcohol was performed using Al2O3 prepared by solvothermal synthesis for selective production of long-chain linear-alpha-olefins (LAO). The effect of the synthesis temperature of alumina precursors on the dehydration of 1-octadecanol (C18H38O) was examined based on the textural properties and Lewis acid-base properties of the catalysts. Amorphous alumina synthesized at 325 °C showed the highest surface area (233.07 m2/g) and total pore volume (1.237 cm3/g) among the catalysts and the best dehydration results: 93% conversion, 62% selectivity of 1-octadecene (C18H36), and 89% LAO purity. This was attributed to the increased Al/O ratio and atomic concentration of surface O in alumina, which were important factors in the catalytic dehydration of 1-octadecanol through the synergistic catalysis of acid-base pairs. The produced bio-based LAO can be key intermediates for synthesis of oxo alcohols and poly-alpha-olefins, as alternatives to petroleum-based LAO to achieve carbon neutrality in chemical industry.

PEPSI++: Fast and Lightweight Network for Image Inpainting.

Among the various generative adversarial network (GAN)-based image inpainting methods, a coarse-to-fine network with a contextual attention module (CAM) has shown remarkable performance. However, due to two stacked generative networks, the coarse-to-fine network needs numerous computational resources, such as convolution operations and network parameters, which result in low speed. To address this problem, we propose a novel network architecture called parallel extended-decoder path for semantic inpainting (PEPSI) network, which aims at reducing the hardware costs and improving the inpainting performance. PEPSI consists of a single shared encoding network and parallel decoding networks called coarse and inpainting paths. The coarse path produces a preliminary inpainting result to train the encoding network for the prediction of features for the CAM. Simultaneously, the inpainting path generates higher inpainting quality using the refined features reconstructed via the CAM. In addition, we propose Diet-PEPSI that significantly reduces the network parameters while maintaining the performance. In Diet-PEPSI, to capture the global contextual information with low hardware costs, we propose novel rate-adaptive dilated convolutional layers that employ the common weights but produce dynamic features depending on the given dilation rates. Extensive experiments comparing the performance with state-of-the-art image inpainting methods demonstrate that both PEPSI and Diet-PEPSI improve the qualitative scores, i.e., the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM), as well as significantly reduce hardware costs, such as computational time and the number of network parameters.

International R&D Collaboration for a Global Aging Society: Focusing on Aging-Related National-Funded Projects.

An international research and development (R&D) collaboration for aging-related projects is necessary to alleviate the severe economic/healthcare/humanitarian challenges of a global aging society. This study presents a practical/systematic framework that enables the provision of information on the research goals, the status of science and technology, and action plans of aging-related program development processes. We used data on aging-related national-funded projects from the United States of America, the European Union, the United Kingdom, Japan, and Korea. We collected data on 6318 national-funded projects, subsequently designating research fields to each project. By analyzing the content of the projects, their representative research fields, and the associated keywords, we assessed the general goals of six different research fields. To recognize the current scientific capabilities of these research fields, we divided the projects by clusters. We provided information on research organizations, specific goals (i.e., project title), project periods, and the funding related to the projects. These may be used by stakeholders in various governments/institutions/industries during future discussions regarding the establishment of an international R&D collaboration strategy. The approach we proposed may facilitate the linkage between knowledge and action during strategy development by maximizing scientific legitimacy, developing consensual knowledge, and minimizing diverging opinions among stakeholders.

Bibliometric analysis of studies about acute myeloid leukemia conducted globally from 1999 to 2018.

A bibliometric study is performed to analyze publication patterns in a specific research area and to establish a landscape model that can be used to quantitatively weigh publications. This study aimed to investigate AML research networks and to conduct a trend-related keyword analysis. We analyzed 48,202 studies about AML published from 1999 to 2019 in the Web of Science Core Collection. The network analysis was conducted using the R&R studio software. The journal Blood had the highest number of published articles with an h-index of 410. The USA had the highest number of total publications (18,719, 38.3%) and research funded by the government, institutions, and pharmaceutical companies (5,436, 10.8%). The institute with the largest number of publications was the MD Anderson Cancer Center. Kantarjian H, Garcia-Manero G, and Ravandi F were the leading authors of publications about AML. Keyword analysis revealed that FLT 3, micro-RNA, and NK cell topics were the hotspots in the cell and gene area in all publications. The overall AML research landscape is popular in the field of translational research as it can identify molecular, cell, and gene studies conducted by different funding agencies, countries, institutions, and author networks. With active funding and support from the Chinese government, the productivity of scientific research is increasing not only in the AML field but also in the medical/health-related science field.

The 100 Most Influential Studies in Chimeric Antigen Receptor T-Cell: A Bibliometric Analysis.

Background: Bibliometric analyses are used to provide information on trends within a specific research field, along with indicators of the impact of a publication. With such an analysis, we map the scientific landscape of chimeric antigen receptor T-cell (CAR-T) research to see the emerging topics and infer directions the field might take. Methods: We extracted the 100 most-cited articles, published all periods (from 2008 to 2019) by the Web of Science Core Collection. Using their bibliographic details, including year of publication, country of author, research organization, author information, and keywords, we graph the networks created between the articles. Results: Of the 100 papers identified, the majority (93%) were written in the USA. Notable was that 34 papers were published from the University of Pennsylvania. Regarding authors, Carl H. June participated in 29 researches, followed by Bruce L. Levine who participated in 12. As for journals, Blood (n = 19) published the most papers, followed by Science Translational Medicine (n = 9) and Cancer Research (n = 9). Lastly, the most frequently used keywords were "adoptive immunotherapy" (n = 47), "lymphocytes" (n = 27), and "antitumor activity" (n = 22). Conclusion: By evaluating the top 100 most-cited papers in the CAR-T field, this study provides insight into the direction of the scientific growth and its trends, as well as information on the field's network structure.

Predominance of emm4 and antibiotic resistance of Streptococcus pyogenes in acute pharyngitis in a southern region of Korea.

BACKGROUND: Streptococcus pyogenes is the most common cause of bacterial pharyngitis. Genotyping of emm is useful for molecular epidemiological survey of S. pyogenes. Antibiotic resistance data are needed for empirical treatments. METHODS: In total, 358 children in Changwon, Korea who had pharyngitis symptoms were subjected to throat cultures to isolate S. pyogenes in 2017. emm genotyping was performed by direct sequencing. An antibiotic susceptibility test was performed using the disk diffusion method for erythromycin (ERY), clindamycin (CLI), tetracycline (TET) and ofloxacin (OFX). Screening for macrolide resistance phenotype and its determinants was performed for the ERY-resistant strains. RESULTS: A total of 190 strains (53.1 %) of S. pyogenes were isolated from 358 children. The most frequent emm genotype was emm4 (53.2 %), followed by emm89 (12.6 %), emm28 (11.6 %) and emm1 (10 %). Antibiotic resistance rates to ERY, CLI, TET and OFX were 3.2 %, 2.6 %, 1.1 % and 2.6%, respectively. There were five isolates of the cMLSB phenotype having the ermB gene and one M phenotype harbouring the mefA gene. CONCLUSIONS: The distribution of emm genotypes was quite different from those previously reported in Korea. emm4 accounted for more than 50  % of the genotypes. Macrolide resistance rates remained very low, but five of six ERY-resistant strains displayed the cMLSB phenotype.

Curcumin and Curcuma longa L. extract ameliorate lipid accumulation through the regulation of the endoplasmic reticulum redox and ER stress.

For this study, we examined the effects of curcumin against acute and chronic stress, paying specific attention to ROS. We also aimed to clarify the differences between acute and chronic stress conditions. We investigated the effects of curcumin against acute stress (once/1 day CCl4 treatment) and chronic-stress (every other day/4week CCl4 treatment). Compared with acute stress, in which the antioxidant system functioned properly and aspartate transaminase (AST) and ROS production increased, chronic stress increased AST, alanine aminotransferase (ALT), hepatic enzymes, and ROS more significantly, and the antioxidant system became impaired. We also found that ER-originated ROS accumulated in the chronic model, another difference between the two conditions. ER stress was induced consistently, and oxidative intra-ER protein folding status, representatively PDI, was impaired, especially in chronic stress. The PDI-associated client protein hepatic apoB accumulated with the PDI-binding status in chronic stress, and curcumin recovered the altered ER folding status, regulating ER stress and the resultant hepatic dyslipidemia. Throughout this study, curcumin and curcumin-rich Curcuma longa L. extract promoted recovery from CCl4-induced hepatic toxicity in both stress conditions. For both stress-associated hepatic dyslipidemia, curcumin and Curcuma longa L. extract might be recommendable to recover liver activity.

Churn prediction of mobile and online casual games using play log data.

Internet-connected devices, especially mobile devices such as smartphones, have become widely accessible in the past decade. Interaction with such devices has evolved into frequent and short-duration usage, and this phenomenon has resulted in a pervasive popularity of casual games in the game sector. On the other hand, development of casual games has become easier than ever as a result of the advancement of development tools. With the resulting fierce competition, now both acquisition and retention of users are the prime concerns in the field. In this study, we focus on churn prediction of mobile and online casual games. While churn prediction and analysis can provide important insights and action cues on retention, its application using play log data has been primitive or very limited in the casual game area. Most of the existing methods cannot be applied to casual games because casual game players tend to churn very quickly and they do not pay periodic subscription fees. Therefore, we focus on the new players and formally define churn using observation period (OP) and churn prediction period (CP). Using the definition, we develop a standard churn analysis process for casual games. We cover essential topics such as pre-processing of raw data, feature engineering including feature analysis, churn prediction modeling using traditional machine learning algorithms (logistic regression, gradient boosting, and random forests) and two deep learning algorithms (CNN and LSTM), and sensitivity analysis for OP and CP. Play log data of three different casual games are considered by analyzing a total of 193,443 unique player records and 10,874,958 play log records. While the analysis results provide useful insights, the overall results indicate that a small number of well-chosen features used as performance metrics might be sufficient for making important action decisions and that OP and CP should be properly chosen depending on the analysis goal.

Protective effect of Curcuma longa L. extract on CCl4-induced acute hepatic stress.

BACKGROUND: The Curcuma longa L. (CLL) rhizome has long been used to treat patients with hepatic dysfunction. CLL is a member of the ginger family of spices that are widely used in China, India, and Japan, and is a common spice, coloring, flavoring, and traditional medicine. This study was performed to evaluate the hepatoprotective activity of CLL extract and its active component curcumin in an acute carbon tetrachloride (CCl4)-induced liver stress model. METHODS: Acute hepatic stress was induced by a single intraperitoneal injection of CCl4 (0.1 ml/kg body weight) in rats. CLL extract was administered once a day for 3 days at three dose levels (100, 200, and 300 mg/kg/day) and curcumin was administered once a day at the 200 mg/kg/day. We performed alanine transaminase (ALT) and aspartate transaminase (AST). activity analysis and also measured total lipid, triglyceride, and cholesterol levels, and lipid peroxidation. RESULTS: At 100 g CLL, the curcuminoid components curcumin (901.63 ± 5.37 mg/100 g), bis-demethoxycurcumin (108.28 ± 2.89 mg/100 g), and demethoxycurcumin (234.85 ± 1.85 mg/100 g) were quantified through high liquid chromatography analysis. In CCl4-treated rats, serum AST and ALT levels increased 2.1- and 1.2-fold compared with the control. AST but not ALT elevation induced by CCl4 was significantly alleviated in CLL- and curcumin-treated rats. Peroxidation of membrane lipids in the liver was significantly prevented by CLL (100, 200, and 300 mg/kg/day) on tissue lipid peroxidation assay and immunostaining with anti-4HNE antibody. We found that CLL extract and curcumin exhibited significant protection against liver injury by improving hepatic superoxide dismutase (p < 0.05) and glutathione peroxidase activity, and glutathione content in the CCl4-treated group (p < 0.05), leading to a reduced lipid peroxidase level. CONCLUSION: Our data suggested that CLL extract and curcumin protect the liver from acute CCl4-induced injury in a rodent model by suppressing hepatic oxidative stress. Therefore, CLL extract and curcumin are potential therapeutic antioxidant agents against acute hepatotoxicity.

Turmeric extract and its active compound, curcumin, protect against chronic CCl4-induced liver damage by enhancing antioxidation.

BACKGROUND: Curcumin, a major active component of turmeric, has previously been reported to alleviate liver damage. Here, we investigated the mechanism by which turmeric and curcumin protect the liver against carbon tetrachloride (CCl4)-induced injury in rats. We hypothesized that turmeric extract and curcumin protect the liver from CCl4-induced liver injury by reducing oxidative stress, inhibiting lipid peroxidation, and increasing glutathione peroxidase activation. METHODS: Chronic hepatic stress was induced by a single intraperitoneal injection of CCl4 (0.1 ml/kg body weight) into rats. Turmeric extracts and curcumin were administered once a day for 4 weeks at three dose levels (100, 200, and 300 mg/kg/day). We performed ALT and AST also measured of total lipid, triglyceride, cholesterol levels, and lipid peroxidation. RESULT: We found that turmeric extract and curcumin significantly protect against liver injury by decreasing the activities of serum aspartate aminotransferase and alanine aminotransferase and by improving the hepatic glutathione content, leading to a reduced level of lipid peroxidase. CONCLUSIONS: Our data suggest that turmeric extract and curcumin protect the liver from chronic CCl4-induced injury in rats by suppressing hepatic oxidative stress. Therefore, turmeric extract and curcumin are potential therapeutic antioxidant agents for the treatment of hepatic disease.

Optimization of lactulose synthesis from whey lactose by immobilized ß-galactosidase and glucose isomerase.

In the present study, commercially available whey was used as a lactose source, and immobilized ß-galactosidase and glucose isomerase were used to synthesize lactulose from whey lactose in the absence of fructose. Optimal reaction conditions, such as lactose concentration, temperature, ionic strength of the buffer, and ratio of immobilized enzymes, were determined to improve lactulose synthesis using immobilized enzymes. Lactulose synthesis using immobilized enzymes improved markedly after optimizing the reaction conditions. When the lactulose synthesis was carried out at 53.5°C using 20% (w/v) whey lactose, 12U/ml of immobilized ß-galactosidase and 60U/ml of immobilized glucose isomerase in 100mM sodium phosphate buffer at pH 7.5, the lactulose concentration and specific productivity were 7.68g/l and 0.32mg/Uh, respectively. Additionally, when the immobilized enzymes were reused for lactulose synthesis, their catalytic activity was 57.1% after 7 repeated uses.

Synthesis of monoclinic potassium niobate nanowires that are stable at room temperature.

We report the synthesis of KNbO(3) nanowires (NWs) with a monoclinic phase, a phase not observed in bulk KNbO(3) materials. The monoclinic NWs can be synthesized via a hydrothermal method using metallic Nb as a precursor. The NWs are metastable, and thermal treatment at ∼450 °C changed the monoclinic phase into the orthorhombic phase, which is the most stable phase of KNbO(3) at room temperature. Furthermore, we fabricated energy-harvesting nanogenerators by vertically aligning the NWs on SrTiO(3) substrates. The monoclinic NWs showed significantly better energy conversion characteristics than orthorhombic NWs. Moreover, the frequency-doubling efficiency of the monoclinic NWs was ∼3 times higher than that of orthorhombic NWs. This work may contribute to the synthesis of materials with new crystalline structures and hence improve the properties of the materials for various applications.

Synthesis of lead-free (K, Na)(Nb, Ta)O3 nanopowders using a sol-gel process.

Lead-free (K0.5Na0.5)(Nb0.7Ta0.3)O3 piezoelectric material was successfully synthesized via a sol-gel process. Crystalline (K0.5Na0.5)(Nb0.7Ta0.3)O3 nanopowders were obtained after heat treatment at 700 degrees C. The particle size was estimated to be 87nm +/- 23 nm. The transmission electron microscopy images showed that individual nanoparticles were single crystalline and had a pseudo-cubic structure with a lattice parameter of -3.96 angstroms. Both X-ray diffraction and scanning electron microscopy studies consistently showed that the crystallization of the (K0.5Na0.5)(Nb0.7Ta0.3)O3 occurred slightly above 500 degrees C. The samples have an appropriate stoichiometry as found via energy dispersive X-ray spectroscopy. The demonstration of the synthesis of (K0.5Na0.5)(Nb0.7Ta0.3)O3 via a sol-gel process as presented in this paper can provide an important foundation for the development of a synthetic route towards (K0.5Na0.5)(Nb0.7Ta0.3)O3 doped with various other elements for high performance piezoelectric devices.

Resistance random access memory based on a thin film of CdS nanocrystals prepared via colloidal synthesis.

We demonstrate that resistance random access memory (RRAM) can be fabricated based on CdS-nanocrystal thin films. A simple drop-drying of the CdS-nanocrystal solution leads to the formation of uniform thin films with controlled thickness. RRAMs with a Ag/Al(2) O(3) /CdS/Pt structure show bipolar switching behavior, with average values of the set voltage (V(Set) ) and reset voltage (V(Reset) ) of 0.15 V and -0.19 V, respectively. The RRAM characteristics are critically influenced by the thickness of the Al(2) O(3) barrier layer, which prevents significant migration of Ag into the CdS layer as revealed by Auger electron spectroscopy (AES). Interestingly, RRAM without an Al(2) O(3) layer (i.e., Ag/CdS/Pt structure) also shows bipolar switching behavior, but the polarity is opposite to that of RRAM with the Al(2) O(3) layer (i.e., Ag/Al(2) O(3) /CdS/Pt structure). The operation of both kinds of devices can be explained by the conventional conductive bridging mechanism. Additionally, we fabricated RRAM devices on Kapton film for potential applications in flexible electronics, and the performance of this RRAM device was comparable to that of RRAMs fabricated on hard silicon substrates. Our results show a new possibility of using chalcogenide nanocrystals for RRAM applications.

Construction of healthy arteries using computed tomography and virtual histology intravascular ultrasound.

Vessel geometry for numerical analysis is generally obtained by computed tomography (CT) or magnetic resonance imaging (MRI) and intravascular ultrasound (IVUS). Most medical imaging is obtained from patients for hemodynamic analysis due to the properties of vascular disease and the difficulties in angiography. To predict the site where plaque occurs and understand the progression of the lesion, however, it is necessary to take into consideration not only the diseased artery, but also the blood flow characteristics of healthy artery. In order to simulate healthy vessels prior to lesion formation, we performed CT and virtual histology intravascular ultrasound (VH-IVUS) on three actual patients and this data was used to develop criteria for healthy vessel construction, a method that virtually removes all intravascular plaque. The lumen of a vessel generated by CT and the lumen from VH-IVUS were compared, and the cross-sectional areas of plaque components (fibrous, fibrofatty, dense calcium, and necrotic) and the lumen from VH-IVUS were analyzed. Geometric differences in the healthy vessel and diseased vessel were analyzed, and flow characteristics of the healthy vessel and diseased vessel were compared through computational fluid dynamics simulation. Low average wall shear stress (AWSS) was distributed in the site where plaque was removed from the healthy vessel, and a high oscillatory shear index (OSI) was observed in the region proximal to the site where plaque previously existed. Low AWSS and high OSI are widely accepted indicators of plaque formation or the direction of plaque progression. A numerical model that effectively predicts lesion forming sites was also generated based on the healthy vessel construction method presented in this study.

Cellulosic ethanol production on temperature-shift simultaneous saccharification and fermentation using the thermostable yeast Kluyveromyces marxianus CHY1612.

In cellulosic ethanol production, use of simultaneous saccharification and fermentation (SSF) has been suggested as the favorable strategy to reduce process costs. Although SSF has many advantages, a significant discrepancy still exists between the appropriate temperature for saccharification (45-50 °C) and fermentation (30-35 °C). In the present study, the potential of temperature-shift as a tool for SSF optimization for bioethanol production from cellulosic biomass was examined. Cellulosic ethanol production of the temperature-shift SSF (TS-SSF) from 16 w/v% biomass increased from 22.2 g/L to 34.3 g/L following a temperature shift from 45 to 35 °C compared with the constant temperature of 45 °C. The glucose conversion yield and ethanol production yield in the TS-SSF were 89.3% and 90.6%, respectively. At higher biomass loading (18 w/v%), ethanol production increased to 40.2 g/L with temperature-shift time within 24 h. These results demonstrated that the temperature-shift process enhances the saccharification ratio and the ethanol production yield in SSF, and the temperature-shift time for TS-SSF process can be changed according to the fermentation condition within 24 h.