Learning to Discretize: Solving 1D Scalar Conservation Laws via Deep Reinforcement Learning

Conservation laws are considered to be fundamental laws of nature. It has broad applications in many fields, including physics, chemistry, biology, geology, and engineering. Solving the differential equations associated with conservation laws is a major branch in computational mathematics. The recent success of machine learning, especially deep learning in areas such as computer vision and natural language processing, has attracted a lot of attention from the community of computational mathematics and inspired many intriguing works in combining machine learning with traditional methods. In this paper, we are the first to view numerical PDE solvers as an MDP and to use (deep) RL to learn new solvers. As proof of concept, we focus on 1-dimensional scalar conservation laws. We deploy the machinery of deep reinforcement learning to train a policy network that can decide on how the numerical solutions should be approximated in a sequential and spatial-temporal adaptive manner. We will show that the problem of solving conservation laws can be naturally viewed as a sequential decision-making process, and the numerical schemes learned in such a way can easily enforce long-term accuracy. Furthermore, the learned policy network is carefully designed to determine a good local discrete approximation based on the current state of the solution, which essentially makes the proposed method a meta-learning approach. In other words, the proposed method is capable of learning how to discretize for a given situation mimicking human experts. Finally, we will provide details on how the policy network is trained, how well it performs compared with some state-of-the-art numerical solvers such as WENO schemes, and supervised learning based approach L3D and PINN, and how well it generalizes.     

补骨脂素抗增生性瘢痕作用机制研究

目的探讨补骨脂素抗增生性瘢痕的作用机制。方法体外培养成纤维细胞,按随机数字表法分为正常组(培养正常成纤维细胞)、瘢痕组(培养增生性瘢痕成纤维细胞)、TGF-β1组(10 ng/ml TGF-β1处理增生性瘢痕成纤维细胞5 min^12 h)、Smurf2 RNA干扰组[Smad泛素化调节因子2(Smad ubiquitin regulatory factor2,Smurf2)siRNA转染增生性瘢痕成纤维细胞72 h]、补骨脂素组(10μmol/L补骨脂素处理增生性瘢痕成纤维细胞继续培养72 h)、补骨脂素+TGF-β1组(增生性瘢痕成纤维细胞加入补骨脂素培养72 h后加入TGF-β1培养6 h)。采用Western blot法检测Smurf2、α-平滑肌肌动蛋白(α-actin SMA,α-SMA)蛋白表达;RT-PCR法检测Ⅰ型胶原蛋白mRNA表达;ELISA法检测TGF-β1蛋白分泌。结果与正常组比较,瘢痕组Smurf2蛋白[(0.83±0.08)比(0.38±0.07)]表达增加(P<0.05);与瘢痕组比较,Smurf2 RNA干扰组TGF-β1[(2.2±0.18)比(4.2±0.47)]表达降低(P<0.05);TGF-β1组Smurf2[(0.71±0.06)比(0.42±0.04)]、α-SMA[(1.42±0.12)比(0.91±0.09)]蛋白表达增加(P<0.05),Ⅰ型胶原蛋白mRNA[(0.72±0.09)比(0.41±0.07)]表达增加(P<0.05);补骨脂素组Smurf2[(0.05±0.01)比(0.42±0.04)]、α-SMA[(0.71±0.07)比(0.91±0.09)]蛋白表达降低(P<0.05),Ⅰ型胶原蛋白mRNA表达[(0.12±0.04)比(0.41±0.07)]降低(P<0.05)。结论补骨脂素可能通过TGF-β1/Smurf2信号通路抑制α-SMA蛋白表达,从而降低Ⅰ型胶原蛋白表达,起到抑制瘢痕形成的作用。     

改良俯卧位对俯卧位通气临床疗效及并发症的影响

目的观察改良俯卧位对俯卧位通气(PPV)临床疗效及并发症的影响。方法将该院呼吸和危重症医学科行PPV的52例患者随机分为对照组和实验组各26例,对照组应用常规方法进行PPV,实验组采用改良俯卧位进行PPV。比较两组患者PPV的临床效果和并发症发生情况。结果两组间俯卧位前和俯卧位后12 h的氧合指数无明显差异(P>0.05);两组患者俯卧位后12 h的氧合指数均较各组俯卧位前明显改善(P<0.05);两组患者气管插管脱出、血流动力学显著波动、误吸和面部水肿的发生率无显著差异(P>0.05);实验组皮肤压疮的发生率明显低于对照组(P<0.05)。结论改良俯卧位对PPV的临床疗效无显著影响,但可以降低皮肤压疮的发生率。     

Analysis of Risk and Burden of Dysentery Associated with Floods from 2004 to 2010 in Nanning,China

This study aimed to examine the association between floods and the morbidity of dysentery and to quantify the burden of dysentery due to floods in Nanning, China. A generalized additive mixed model was conducted to assess the relationship between monthly morbidity of dysentery and floods from 2004 to 2010. The years lived with disability (YLDs) of dysentery attributable to floods were then estimated based on the WHO framework of the burden of disease study for calculating the potential impact fraction. The relative risk (RR) of floods on the morbidity of dysentery was 1.44 (95% confidence interval [CI] = 1.18–1.75). The models suggest that a potential 1-day rise in flood duration may lead to 8% (RR = 1.08, 95% CI = 1.04–1.12) increase in the morbidity of dysentery. The average attributable YLD per 1,000 of dysentery caused by floods were 0.013 in males, 0.005 in females, and 0.009 in persons. Our study confirms that floods have significantly increased the risk and the burden of dysentery in the study area. Public health action should be taken to prevent and control the potential risk of dysentery after floods. Vulnerable groups such as males and children should be paid more attention.     

Evidence of impaired pain modulation in adolescents with idiopathic scoliosis and chronic back pain

BACKGROUND CONTEXT

Although 40% of adolescent idiopathic scoliosis (AIS) patients present with chronic back pain, the pathophysiology and underlying pain mechanisms remain poorly understood. We hypothesized that development of chronic pain syndrome in AIS is associated with alterations in pain modulatory mechanisms.

“医教研”三轴联动培养医学检验技术专业应用型人才的改革思路与实践

医学检验技术隶属医学技术一级学科，注重技术型人才培养。该专业教育目前存在重理论知识教学、轻实践技能培训，部分课程实用性不强，毕业生专业素质欠缺等问题。本研究将“教学”“临床”“科研”三轴结合，通过完善师资队伍，增加实验课程、临床技能和科研实训比例；依托“验证→综合→设计”实验、实验室开放和本科生科研竞赛等抓手，开展专业人才培养工作；取得较好教育成效，对检验学科发展也有一定助益。     

新型雌激素受体β信号通路调节因子的分离及鉴定

目的：利用酵母双杂交技术筛选与ERβ相互作用的蛋白，为进一步研究ERβ的功能奠定基础。方法：从乳腺cDNA文库中钓取ERβ-AF1 cDNA，构建诱饵蛋白载体pAS2-1-ERβ—AF1，利用酵母双杂交筛选技术，在乳腺cDNA文库中筛选与ERβ-AF1相互作用的蛋白。将含有pACT2-候选基因的酵母菌与含有pAS2—1-ERp—AF1的酵母菌进行交配实验，利用LacZ报告基因检测ERβ—AF1与候选蛋白之间是否存在相互作用。然后将ERp—AF1和候选蛋白分别构建到pGEx-KG和pET-28a上，并纯化GST-ERβ-AF1和His-侯选蛋白融合蛋白，利用GST沉淀实验进一步验证ERβ-AF1和候选蛋白的体外相互作用。结果：所获取的候选蛋白为CTGF，是CCN多肽家族成员。交配实验表明，CTGF与ERβ-AF1特异相互作用，而不与空载体或BRCT1对照相互作用。GST沉淀实验进一步验证，CTGF与ERβ、ERα均存在相互作用，但与CTGF同源性较高的NOV蛋白却不与ERβ、ERa结合。结论：CTGF和ERβ存在特异的相互作用。