基于视触觉与深度学习的猕猴桃无损硬度检测方法

硬度是确定猕猴桃成熟度的重要指标之一,对其贮藏周期与销售节点均具有重要指导意义。针对现阶段缺乏使用简易、成本低且精度高的猕猴桃无损硬度检测方法的问题,提出了一种基于视触觉与深度学习的猕猴桃硬度检测方法,通过分析柔性触觉传感层与猕猴桃接触时的形变,获取猕猴桃的动态触觉信息,并据此推断其硬度。以树莓派开发板为机电控制平台,制作了猕猴桃视触觉序列图像采集装置,并对装置按压猕猴桃间隔3h后接触面果肉与非接触面果肉的CIELAB颜色分量平均数进行差异显著性检验,随后采集了猕猴桃视触觉序列图像数据集600组,分别搭建了CNN网络、CNN-LSTM迁移学习网络、CNN-LSTM联合学习网络对视触觉序列图像进行分析及硬度预测。研究结果表明,接触面果肉与非接触面果肉颜色L*、a*、b*三通道分量下平均值无显著差异;深度学习模型LSTM引入长时和短时信息可以动态关联CNN提取的单帧图像特征,从而有效推断猕猴桃硬度,其中CNN-LSTM联合学习模型预测效果最优,其均方根误差(RMSE)、平均绝对误差(MAE)、决定系数R2分别为 1.611N、1.360N、0.856,优于现阶段光谱技术检测猕猴桃硬度的结果,随后将模型嵌入树莓派中制作了猕猴桃硬度自动检测装置,可实现短时间内猕猴桃硬度的较为准确检测。因此,结合视触觉传感方法与联合学习模型可以实现对单个猕猴桃硬度的准确无损测量。     

In vitro sensitivity of the amphibian pathogen Batrachochytrium dendrobatidis to antifungal therapeutics

Chytridiomycosis, a skin disease caused by Batrachochytrium dendrobatidis, has caused amphibian declines worldwide. Amphibians can be treated by percutaneous application of antimicrobials, but knowledge of in vitro susceptibility is lacking. Using a modified broth microdilution method, we describe the in vitro sensitivity of two Australian isolates of B. dendrobatidis to six antimicrobial agents. Growth inhibition was observed, by measurement of optical density, with all agents. Minimum inhibitory concentrations (µg/ml; isolate 1/2) were – voriconazole 0.016/0.008; itraconazole 0.032/0.016; terbinafine 0.063/0.063; fluconazole 0.31/0.31; chloramphenicol 12.5/12.5; amphotericin B 12.5/6.25. Killing effects on zoospores were assessed by observing motility. Amphotericin B and terbinafine killed zoospores within 5 and 30 min depending on concentration, but other antimicrobials were not effective at the highest concentrations tested (100 µg/ml). This knowledge will help in drug selection and treatment optimization. As terbinafine was potent and has rapid effects, study of its pharmacokinetics, safety and efficacy is recommended.     

Calorific values and ash contents of different parts of Masson pine trees in southern China

Calorific value of plants is an important parameter for evalu- ating and indexing material cycles and energy conversion in forest eco- systems. Based on mensuration data of 150 sample sets, we analyzed the calorific value （CV） and ash content （AC） of different parts of Masson pine （Pinus massoniana） trees in southern China using hypothesis testing and regression analysis. CV and AC of different tree parts were almost significantly different （P〈0.05）. In descending order, ash-free calorific value （AFCV） ranked as foliage 〉 branch 〉 stem bark 〉 root 〉 stem wood, and AC ranked as foliage 〉 stem bark 〉 root 〉 branch 〉 stem wood. CV and AC of stem wood from the top, middle and lower sections of trees differed significantly. CV increased from the top to the lower sections of the tnmk while AC decreased. Mean gross calorific value （GCV） and AFCV of aboveground parts were significantly higher than those of belowground parts （roots）. The mean GCV, AFCV and AC of a whole tree of Masson pine were 21.54 kJ/g, 21.74 kJ/g and 0.90%, re- spectively. CV and AC of different tree parts were, to some extent, cor- related with tree diameter, height and origin.     

Genetic variation and selection of introduced provenances of Siberian Pine （Pinus sibirica） in frigid regions of the Greater Xing’an Range,Northeast China

Siberian Pine （Pinus sibirica） is an ecologically and eco-nomically important species in pristine forests throughout northern Rus-sia. Four provenances of P. sibirica were introduced from Mongolia and Russia to the Greater Xing’an Range （the Daxing’anling）, northeast China in 1993. The aim of this research was to study genetic variation and selection of the introduced four Pinus sibirica provenances. Heights （H）, basal diameters （BD）, survival rates （SR） and crown lengths （CL） of different families were measured as primary outcomes in different growth years. Results of data analyses demonstrated high coefficients of phenotypic variation （PCV） and heritability （H2） for H, BD and CL at 18 years after introduction. PCV and H2 increased with age. Correlations of＆amp;nbsp;growth traits between any two growth years were all significantly positive, but the correlation coefficient was smaller when the growth year interval was larger. Correlations between H and the original environment factors decreased gradually, indicating that with long-term subsistence in the new environment, the influence of the source environment declined. Colligation of multiple traits to estimate provenances showed that Novosibirsk, Tomsk, and Altai Mountains had higher survival rates and biomass, and proved more suitable for introduction and plantation in the Greater Xing’an Range in China.     

Feasibility and repeatability of cold and mechanical quantitative sensory testing in normal dogs

Feasibility and inter-session repeatability of cold and mechanical quantitative sensory testing (QST) were assessed in 24 normal dogs. Cold thermal latencies were evaluated using a thermal probe (0 °C) applied to three pelvic limb sites. Mechanical thresholds were measured using an electronic von Frey anesthesiometer (EVF) and a blunt-probed pressure algometer (PA) applied to the dorsal aspect of the metatarsus. All QST trials were performed with dogs in lateral recumbency. Collection of cold QST data was easy (feasible) in 19/24 (79%) dogs. However, only 18.4%, 18.9% and 13.2% of cold QST trials elicited a response at the medial tibia, third digital pad and plantar metatarsal regions, respectively. Collection of mechanical QST data was easy (feasible) in 20/24 (83%) dogs for both EVF and PA.At consecutive sampling times, approximately 2 weeks apart, the average EVF sensory thresholds were 414 ± 186 g and 379 ± 166 g, respectively, and the average PA sensory thresholds were 1089 ± 414 g and 1028 ± 331 g, respectively. There was no significant difference in inter-session or inter-limb threshold values for either mechanical QST device. The cold QST protocol in this study was achievable, but did not provide consistently quantifiable results. Both mechanical QST devices tested provided repeatable, reliable sensory threshold measurements in normal, client-owned dogs. These findings contribute to the validation of the EVF and PA as tools to obtain repeated QST data over time in dogs to assess somatosensory processing changes.     

Sensitive detection of the c‐KIT c.1430G>T mutation by mutant‐specific polymerase chain reaction in feline mast cell tumours

Here, we describe the establishment of mutant‐specific polymerase chain reaction (PCR) for detection of a c‐KIT c.1430G>T mutation in feline mast cell tumours. Several mutations in feline c‐KIT have been identified, with the c.1430G>T mutation accounting for a significant portion of feline mast cell tumour mutations. The c.1430G>T mutation in c‐KIT exon 9 was detected in 15.7% (11 of 70) of samples by mutant‐specific PCR but in only 7.1% (5 of 70) by PCR–restriction fragment length polymorphism (RFLP) in the genomic DNA isolated from 70 formalin‐fixed paraffin‐embedded sections or cells collected by fine needle aspiration. Mutant‐specific PCR showed remarkably higher detection rate than did PCR–RFLP. DNA sequence analysis did not always yield identical results to those of mutant‐specific PCR, suggesting heterogeneity of tumour cells. Mutant‐specific PCR is a valid and efficient screening tool for detection of the c‐KIT c.1430G>T point mutation in feline mast cell tumours compared with PCR–RFLP and sequencing analysis.     

The influence of various confounding factors on mechanical nociceptive thresholds in the donkey

ObjectiveTo evaluate a mechanical nociceptive threshold (MNT) testing device in the donkey, and to investigate the influence of potential confounders on MNTs generated.Study designProspective, randomised.AnimalsSixteen castrated male donkeys aged 4–9 years, weighing 105–170 kg.MethodsMechanical nociceptive thresholds were measured using an actuator with three pins placed on the dorsal aspect of the distal limb, connected to a force meter. The pins (surface area 15 mm²) were extruded onto the limb by pressurising an air-filled syringe, until the MNT force (when foot-lift was observed) or 25 N (cut-off force) was reached. Effect on MNT of presence of a companion donkey, the limb tested, rate of application of force, testing location, level of distraction, ambient temperature and hair cover at the test site was evaluated. Long and short-term repeatability of MNT was assessed. Data were analysed using general linear models and Mann–Whitney U tests, p < 0.05 was considered significant.ResultsIncreasing the rate of force application significantly increased the mean ± SD MNT from 9.2 ± 2.0 N when applied at 0.4 N sec^?1 to 10.6 ± 2.1 N when applied at 1.2 N sec^?1 (p = 0.001). No other factors significantly influenced MNT. Mean MNT remained stable over a 3 week period, however MNTs were significantly (p = 0.006) higher (12.8 ± 3.0 N cf 10.3 ± 1.9 N) after a 12 month interval.Conclusions and clinical relevanceWhen designing studies measuring MNT in donkeys, rate of application of force must be standardised. Donkeys’ MNTs have good short-term stability suggesting this technique is appropriate for short-term analgesiometry studies; however variability of MNTs over the long-term is greater.     

Growth testing of two breeds of common carp, Cyprinus carpio L. (Hungarian mirror and Třeboň scaly carp), in ponds with low and high stocking density

The growth of two breeds of common carp, Cyprinus carpio L., was tested in ponds under the climatic conditions of South Bohemia. T?eboň scaly carp (TR) and Hungarian mirror carp (M2) were kept in both low and high stocking densities during the second growing season and then stocked together for communal testing during the third growing season. Before the communal testing, the mean initial weights of fish from low‐ and high‐density stocks differed significantly (374.1 vs. 227.7 g for the TR breed and 766.7 vs. 317.3 g for the M2 breed respectively, P<0.01). After communal testing, mean weights of fish from low‐ and high‐density stocks gained 761.8 vs. 543.8 g for the TR breed and 1339.7 vs. 706.7 g for the M2 breed respectively. These observed weights were also significantly different (P<0.01). However, the test of corrected weight gain, i.e. gain not related to the initial weight of fish, revealed insignificant differences (P>0.01) between the weight gains after correction, i.e. the effect of different initial weights was successfully eliminated. These results seem to confirm the applicability of this method for the assessment of growth of purebred common carp under the climatic conditions of Central European fish farms.     

华阴市小麦“3414”试验研究

[目的]为了提出适宜华阴市小麦配方施肥方案.[方法]对小麦“3414”肥效试验结果进行分析,探索氮磷钾不同与小麦产量的关系.[结果]确定适宜华阴市小麦最高产量和最佳产量的施肥模式和目标产量,同时确定单因子肥效效应方程.[结论]该研究可以为华阴市小麦测土配方施肥提供依据.