基于ML-kNN多标记学习的中医体质辨识模型研究

目的 中医体质与人体健康状态密切相关。研究利用人工智能技术辨识中医体质，为中医体质辨识智能化及自动化发展提供新思路。方法 以江苏省中医院体检中心的中医体质数据作为初始数据样本，经过数据清洗、过滤及结构化最终纳入9844条数据作为研究对象，运用ML-kNN多标记k近邻算法构建中医体质辨识模型，使用10折交叉验证训练模型，并采用多标记学习评价指标评估模型效果。结果 中医体质辨识模型的平均汉明损失为0.096 1，平均1-错误率为0.126 1，平均排序损失为0.086 6，平均覆盖率为1.153 5，平均精度为88.57%。结论 基于体检中心中医体质辨识量表数据，利用ML-kNN多标记学习算法，构建体质辨识模型，能够有效实现中医体质辨识智能化。     

Immunotherapy choice and maintenance for generalized myasthenia gravis in China

AimsTo compare long‐term efficacy and safety of immunotherapeutic strategies as maintenance to prevent disease relapses of generalized myasthenia gravis (MG) in real‐world settings.MethodsThis is a retrospective cohort study on generalized MG conducted in seven major neurological centers across China. Eligible participants were patients with generalized MG who were under minimal manifestation status or better. Main outcome measures were probability of patients free of relapses and causes of drug discontinuation.ResultsAmong 1064 patients enrolled, the median (interquartile range) age was 50.3 (37.0‐62.5) years and 641 (60.2%) were women. Disease relapse was significantly lower for rituximab (6.1%) compared with all the other monotherapies (hazard ratio [HR] = 0.18, 95% confidence interval [CI] 0.06 to 0.56, P = .0030). As combination therapies, tacrolimus in combination with corticosteroids reduced risk of disease relapses compared with azathioprine with corticosteroids (HR = 0.45, 95% CI 0.25 to 0.81, P = .0077) or mycophenolate mofetil with corticosteroids (HR = 0.32, 95% CI 0.15 to 0.67, P = .0020). Otherwise, lower‐dose corticosteroids or azathioprine as monotherapy significantly increased risk of disease relapses (HR = 2.78, 95% CI 1.94 to 3.99, P < .0001; HR = 2.14, 95% CI 1.42 to 3.23, P = .0003, respectively). The proportion of discontinuation was lowest in patients with rituximab (20.4%) as monotherapy and tacrolimus with corticosteroids (23.6%). Overall, combination treatment of immunosuppressants with corticosteroids had a lower rate of discontinuation compared with corresponding monotherapy (HR = 0.51, 95% CI 0.36 to 0.71, P < .0001).ConclusionsRituximab as monotherapy and tacrolimus with corticosteroids displayed better clinical efficacy as well as drug maintenance to prevent disease relapses in patients with generalized MG.     

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.     

PI3K抑制剂ZSTK-474对结直肠癌细胞增殖和侵袭的影响

目的:探讨PI3K抑制剂ZSTK-474对结直肠癌细胞增殖和侵袭的影响。方法:采用不同浓度（0、1、2、4、6、8和10 μmol/L）ZSTK-474处理结直肠癌HT-29和HCT-116细胞24 h和48 h，CCK8法检测细胞增殖活性并计算出ZSTK-474在两株结直肠癌细胞中的半数抑制浓度（IC50）。根据IC50的值选用浓度为4 μmol/L的ZSTK-474作用HT-29和HCT-116细胞24 h，同时设置加入等量的0.1%DMSO为对照。采用平板克隆形成实验检测细胞增殖，流式细胞仪检测细胞凋亡，qRT-PCR检测细胞PI3K、Akt的mRNA水平，Western blot 检测细胞PI3K、Akt的蛋白水平。结果:ZSTK-474可明显抑制HT-29和HCT-116细胞的增殖，且呈时间和剂量依赖性。ZSTK-474可抑制HT-29和HCT-116细胞的侵袭能力，但不影响其凋亡。ZSTK-474处理HT-29和HCT-116细胞24 h后，PI3K和Akt的mRNA和蛋白表达水平均显著下降。结论:ZSTK-474能够抑制结直肠癌细胞增殖，并且能抑制其侵袭。     

ArcCHECK系统在食管癌螺旋断层放疗计划验证中的应用

目的 探讨ArcCHECK系统在食管癌断层治疗旋转照射和固定野照射计划验证中的应用,总结相关经验。方法 对32例不同部位食管癌分别制作Helical旋转照射和Direct固定野照射验证计划,并通过ArcCHECK测量、分析,对比验证结果通过率。分析靶体积与计划验证通过率的相关性。将靶体积较小的治疗验证计划分别放在ArcCHECK模体中心和外周探测点处,分析验证通过率差异。结果 Helical计划验证通过率高于Direct计划(P<0.01),其靶体积与验证通过率的相关系数分别为-0.364和-0.042,P值分别为0.041和0.819。Helical计划采用3%/2mm标准时,高剂量区放在模体中心和外周探测点处测得的通过率不同(P=0.005),后者通过率更高;采用3%/3mm标准时与Direct计划的3%/3m、3%/2mm标准的相近(P均>0.05)。结论 Helical计划验证通过率普遍高于Direct计划,原因可能与ArcCHECK探测器的角度响应以及因更多参考点受到低剂量辐射而未参与计算有关,另外还可能跟Direct计划对断层治疗剂量控制系统要求更高有关。在Helical验证计划中,当采用3%/3mm标准时,靶体积越大,验证时出现较低通过率的可能性增加,但相关系数较低。验证计划的高剂量区位于模体中心或者探测点处都可以实现计划验证,综合考量建议放在模体等中心处。     

基于虚拟现实技术的放疗CT模拟定位远程培训系统研发

目的 研发一种基于虚拟现实技术的放疗CT模拟定位的远程培训系统,探索一种医学培训的新方法。方法 使用3DMax与Maya进行3D建模,Unity3D引擎开发3D虚拟操作及交互系统;Java的SpringMvc架构作为系统后台服务,MySQL作为后台数据库系统;并将用户分为教师和学员两种角色,模式分为教学与考核模式。结果 系统功能涵盖CT模拟定位全过程,主要包括患者信息管理、CT模拟定位机认知、体位固定技术、CT定位扫描、处理突发事件等模块。自2018年投入使用以来,运行稳定,系统浏览量达14 920人次,培训通过率为86.66%。与传统培训相比,培训效率明显提升,并获得一致好评。结论 远程培训系统能有效提升学员的临床实践能力、人文关怀能力,具有良好的自主性、共享性、创新性。目前系统已上线且推广性较强,应用前景广阔。     

Cefoperazone-sulbactam and risk of coagulation disorders or bleeding: a retrospective cohort study

ABSTRACT

Objectives: Limited evidence has suggested that cefoperazone-sulbactam causes coagulation disorders and bleeding.

Methods: The authors conducted a retrospective study to compare patients receiving cefoperazone-sulbactam versus those treated with cefoperazone-tazobactam or ceftazidime. Propensity-score matching was used to explore whether treatment with cefoperazone-sulbactam increased the risk of prothrombin time (PT) prolongation, coagulation disorders, and bleeding, or decreased platelets (PLT).

Results: The cohort included 23,242 patients. Among patients receiving cefoperazone-sulbactam, the risk of PT prolongation, coagulation disorders, decreased PLT, and bleeding was 5.3%, 9.2%, 15.7%, and 4.2%, respectively. Propensity-score matching analyses suggested that cefoperazone-sulbactam increased the risk of PT prolongation (aOR 2.26, 95% CI 1.61–3.18), coagulation disorders (aOR 1.81, 95% CI 1.43–2.30), and decreased PLT (aOR 1.46, 95% CI 1.25–1.72), but not increase bleeding (aOR 1.05, 95% CI 0.79–1.40) compared with ceftazidime. Patients receiving cefoperazone-sulbactam had higher risk of PT prolongation (aOR 1.53, 95% CI 1.11–2.10), coagulation disorders (aOR 1.53, 95% CI 1.21–1.95), but not decreased PLT (aOR 0.93, 95% CI 0.81–1.07) or bleeding (aOR 1.11, 95% CI 0.87–1.42), compared with those receiving cefoperazone-tazobactam.

Conclusion: Cefoperazone-sulbactam may be associated with a higher risk of PT prolongation and coagulation disorders compared with cefoperazone-tazobactam and ceftazidime.     

Regional volume changes of the brain in migraine chronification

The pathophysiology of migraine is complex.Neuroimaging studies reveal functional and structural changes in the brains of migraine patients.We sought to explore regional volume differences in intracranial structures in patients with episodic and chronic migraine.Sixteen episodic migraine patients,16 chronic migraine patients,and 24 normal controls were recruited and underwent 3.0 T MRI scanning.The volumes of 142 brain regions were calculated by an automatic volumetric algorithm and compared with clinical variables.Results demonstrated that the volumes of specific regions in the frontal and occipital lobes,and the right putamen,were increased and the volume of the fourth ventricle was decreased in the episodic migraine patients compared with controls.The volumes of the left basal forebrain,optic chiasm,and,the fourth ventricle were decreased in the chronic migraine patients,while the occipital cortex and the right putamen were larger.Compared to episodic migraine patiants,chronic migraine patients displayed larger left thalamus and smaller frontal regions.Correlation analysis showed that headache frequency was negatively correlated with the volume of the right frontal pole,right lateral orbital gyrus,and medial frontal lobes and positively correlated with the volume of the left thalamus.The sleep disturbance score was negatively correlated with the volume of the left basal forebrain.This suggests that migraine patients have structural changes in regions associated with pain processing and modulation,affective and cognitive processing,and visual perception.The remodeling of selective intracranial structures may be involved in migraine attacks.This study was approved by the Ethics Committee of Chinese PLA General Hospital(approval No.S2018-027-02)on May 31,2018.     

慢性阻塞性肺疾病患者肺癌高危筛查的研究进展

肺癌和慢性阻塞性肺疾病(COPD)是两种密切相关的疾病,对公共卫生有很大影响。每年对胸部进行低剂量计算机断层扫描的筛査可以显著降低肺癌病死率。COPD与肺癌发病风险增加有关。流行病学研究发现COPD患者与无气流阻塞的患者相比,肺癌风险增加2〜4倍。COPD合并肺癌发病风险增加似乎是由于肺气肿引起的。肺气肿已被证明是一个重要的肺癌危险因素。COPD合并肺癌筛査评分以确定COPD合并肺癌发病高危人群,是COPD人群肺癌筛查的一项有益的探索.需要更多临床实践的证明。