Search

This study aimed to determine whether the prevalence of thyroid nodules (TNs) increased due to modern lifestyles or other factors, despite the advances in screening and diagnostic tools.This study included 3474 pairs of participants, who were matched by gender and age (±3 years) from two cross-sectional sampling surveys: (1) the program on the iodine nutritional status and related health status of residents in Shanghai in 2009; (2) the thyroid disease screening program for adults in Shanghai between 2017 and 2018. The prevalence of TNs and thyroid diseases in 2009 and 2017-2018 were compared, and the potential risk factors of TNs were detected.The prevalence of TNs in 2009 was 28.9%: 22.5% in males and 34.5% in females. In 2017, this increased to 43.8%: 37.9% in males and 49.1% in females. The prevalence of TNs significantly increased from 2009 to 2017 (odds ratio, 1.486; 95% confidence interval, 1.238-1.786). In addition, female gender, thyroid disease history, and age were the main risk factors for TNs after adjusting for confounders in the logistic regression across the time period.The prevalence of TNs significantly increased across nearly 10 years in Shanghai.

In the model compression, to achieve the optimal pruning effect, weighted pruning or filter pruning was used to prune the network in traditional methods, in fact, the network can continue to be compressed. Therefore, a hybrid pruning method of convolution kernel weight pruning and filter pruning are proposed in this paper. First of all, the weight pruning is carried out for the convolutional neural network model, this paper adopts the method of dynamic convolution kernel for weight pruning, and the fluctuation value of each time normalized value in network update process is taken as the pruning criterion for the weight of the convolutional kernel. Because the weight pruning of the convolution kernel is dynamic, the floating-point operation (FLOP) is significantly reduced, and the parameter scale does not decrease significantly. Then, the model was pruning by convolution kernel l 1 -norm [1] method, which is not only effectively reduce the parameter scale, but also no extra calculations are introduced. In the case of a small loss of network accuracy, the hybrid pruning of the convolution kernel can maximize the compression of the model to meet the conditions for deployment in mobile edge devices, and has significant improvements in floating-point operations and parameter compression. Therefore, mixed pruning in this paper can effectively make up for the shortage of single pruning method. Through VGG-16, ResNet and other networks, it has been proved in practice that the convolution kernel hybrid pruning method proposed in this paper can effectively reduce the model size and maintain a high level of accuracy.

This study investigated the effects of dietary arginine (Arg) supplementation, just during lactation, on sow and litter performance, plasma concentrations of metabolites and hormones, and milk yield and composition in multiparous sows. Thirty-one sows were randomly assigned to 3 dietary treatments supplemented with 0.0 (control, n=10), 0.5% (n=10), or 1.0% (n=11) L-Arg-HCI, respectively. Experimental diets were provided to the sows from d 3 to 21 of lactation. Plasma and milk samples were collected at d 14 and 21 of lactation. The average daily gain (ADG) of piglets from sows fed diets supplemented with 0.5 or 1.0% L-Arg-HCI at d 3 to 14 of lactation, were higher than that of controls (P

Four series of simple and rodlike racemic biphenyl mesogens possessing a 2-octyloxy tail and different substituents at the inner-core position of the phenyl ring were easily prepared. The mesophases of these racemic biphenyl mesogens were confirmed by variable-temperature XRD and the characteristic texture of POM. In general, cubic BPs can be induced by adding an appropriate ratio of chiral additive S811 or ISO(6OBA)2 into these racemic biphenyl mesogens during the heating and cooling processes. Interestingly, BPIII (5–6 K) easily dominates in highly chiral conditions for the blended mixture composed of mono-substituted biphenyl mesogens and chiral dopant S811. In addition, the formation temperature of BPIII is near room temperature (36 °C) when compound C6OBiPhI-OH is blended with 35 wt% S811 during the cooling process. Stable BPs with more than 20 K present in the blended mixture system were composed of chiral dopant ISO(6OBA)2 and no substituted biphenyl mesogens CnOBiPhI-H or difluoro substituted CnOBiPhI-FF. Notably, the widest temperature range of BP (∼34 K) can be induced by adding only 10 wt% chiral dopant ISO(6OBA)2 with high HTP into the biphenyl compound C6OBiPhI-H. The properties of the BPs were characterized by POM, DSC and reflection spectra. On the basis of our experimental results and molecular modeling, we demonstrated that the appearance and temperature range of BPs in this simple type of biphenyl mesogen with 2-octyloxy tail are affected by the molecular dipole moment and biaxiality.

This study was designed to evaluate the prevalence of thyroid nodules (TNs) and its relationship with urine iodine concentrations (UICs) after the regional rapid economic growth and lifestyle changes.A cross-sectional survey was conducted in the general population aged 15-69 years. A questionnaire regarding general and personal characteristics and relevant information was administered. Ultrasonography of the thyroid was performed, and serum triiodothyronine (T3), tetraiodothyronine (T4), serum thyroid stimulating hormone (TSH), free triiodothyronine (FT3), free tetraiodothyronine (FT4), thyroglobulin antibody (TgAb), thyroid peroxidase antibody (TPOAb), and TSH receptor antibody (TRAb) levels were measured for each individual subject.The prevalence rates of TNs in the whole population, females and males were 27.76%, 34.04%, and 21.60%, respectively. The prevalence of multiple nodules increased with age, whereas the prevalence peaks differed between males and females. The median UICs in the whole population and females with non-TNs were higher than those of subjects with TNs (P=0.0035, P=0.0068). The median UICs in subjects with a single TN were higher than those in subjects with multiple TNs (P=0.0164, P=0.0127). The result showed a U-shaped curve relationship between UIC and prevalence of TNs. The prevalence of TNs was the lowest when the UIC was 140-400 μg/L.The prevalence of TNs was nearly 30% and increased with age. The relationship between UIC and prevalence of TNs is U-shaped, with an increase in risk when the UIC was140 μg/L and400 μg/L. Very low or high UIC levels need attention and correction.