Named entity recognition is a fundamental task in natural language processing. Recognizing entities in biomedical text, known as the BioNER, is particularly crucial for cutting-edge applications. However, BioNER poses greater challenges compared to traditional NER due to (1) nested structures and (2) category correlations inherent in biomedical entities. Recently, various BioNER models have been developed based on region classification or large language models. Despite being successful, these models still struggle to balance handling nested structures and capturing category knowledge. We present a novel parallel BioNER model, Bean, designed to address the unique properties of biomedical entities while achieving a reasonable balance between handling nested structures and incorporating category correlations. Extensive experiments on five public NER datasets, including four biomedical datasets, demonstrate that Bean achieves state-of-the-art performance. The proposed Bean is elaborately designed to achieve two key objectives of the BioNER task: clearly detecting entity boundaries and correctly classifying entity categories. It is the first BioNER model to handle nested structures and category correlations in parallel. We exploit head, tail, and contextualized features to efficiently detect entity boundaries via a triaffine model. To the best of our knowledge, we are the first to introduce a multi-label classification model for the BioNER task to extract entity category information without boundary guidance.

阿尔茨海默病(Alzheimer's disease,AD)给社会带来了巨大的医疗和经济负担,寻找和发现其治疗药物有着重大的研究意义.本研究采用知识图谱嵌入在公开的药物再利用知识图谱(drug repurposing knowledge graph,DRKG)上研究了AD的药物重定位.首先,利用4种知识图谱嵌入模型,即TransE、DistMult、ComplEx和RotatE在DRKG上学习实体和关系的嵌入向量;随后使用3种经典的知识图谱评估指标评估和比较了这些模型的性能和学习到的嵌入向量的质量;根据评估比较的结果,选择利用RotatE模型进行链接预测,确定了16种有可能用于AD治疗的药物,其中谷胱甘肽、氟哌啶醇、辣椒素、槲皮素、雌二醇、葡萄糖、双硫仑、腺苷、帕罗西汀、紫杉醇、格列本脲、阿米替林已被前人的研究证实对于AD有潜在的治疗作用.研究结果表明,基于知识图谱嵌入的药物重定位研究有望为AD药物发现提供新的思路和方法,RotatE模型可以有效地整合DRKG的多源信息,进而很好地完成了AD药物重定位任务.本研究的源代码可以从获得.

知识图谱技术促进了新药研发的进展,但国内研究起点晚且领域知识多以文本形式存储,图谱重用率低.因此,本研究基于多源异构的医药文本,设计了以Bert-wwm-ext预训练模型为基础,并融合级联思想的中文命名实体识别模型,从而减少了传统单次分类的复杂度,进一步提高了文本识别的效率.实验结果显示,该模型在自建的训练语料上的F1分数达0.903,精确率达89.2％,召回率达91.5％.同时,将模型应用于公开数据集CCKS2019上,结果显示该模型能够更好地识别中文文本中的医疗实体.最后,利用此模型构建了一个中文医药知识图谱,图谱包含13 530个实体,10 939个属性,以及39 247个相关关系.本研究所提出的中文医药实体识别与图谱构建方法,有望助力研究者加快医药知识新发现,从而缩短新药研发进程.

Sleep quality is related to people's physical and mental health, so an accurate assessment of sleep quality is key to recognizing sleep disorders and taking effective interventions. To address the shortcomings of traditional manual and automatic staging methods, such as being time-consuming and having low classification accuracy, an automatic sleep staging method based on the null space pursuit (NSP) decomposition algorithm of single-channel electroencephalographic (EEG) signals is proposed, which provides a new way for EEG signal decomposition and automatic identification of sleep stages. First, the single-channel EEG signal data from the Sleep-EDF database, DREAMS Subject database, and Sleep Heart Health Study database (SHHS), available on PhysioNet, were preprocessed, respectively. Second, the preprocessed single-channel EEG signals were decomposed by the NSP algorithm. Third, we extracted nine features in the time domain of the nonlinear dynamics and statistics from the original EEG signal and the six simple signals that were decomposed. Finally, the extreme gradient boosting (XGBOOST) algorithm was used to construct a classification model to classify and identify the 63 extracted EEG signal features for automatic sleep staging. The experimental results showed that, on the Sleep-EDF database, the accuracy of four and five categories were 93.59% and 92.89%, respectively; on the DREAMS Subject database, the accuracy rates of four and five categories were 91.32% and 90.01%, respectively; on the SHHS database, the accuracy rates of four and five categories were 90.25% and 88.37%, respectively. The experimental results show that the automatic sleep staging model proposed in this work has high classification accuracy and efficiency, as well as strong applicability and robustness.

预测药物在血浆中的蛋白结合率,有助于了解药物的药代动力学特征,对药物发现的早期研究有重要的参考价值.本研究收集了 2452个临床药物的血浆蛋白结合率信息,用Molecular Operating Environment(MOE)和Mordred两种软件计算分子描述符,将算得的分子描述符作为模型的输入特征.使用极端梯度提升(extreme gradient boosting,XGBoost)算法和随机森林(randomforest,RF)算法构建机器学习模型.结果表明,与MOE相比,将Mordred计算的分子描述符作为模型的输入,构建的模型预测性能更优.使用XGBoost算法和RF算法构建模型的预测性能结果相近,最优模型的R2均为0.715.此外,根据研究结果得出药物血浆蛋白结合率与药物分子的一些理化性质参数,如水溶性,辛醇/水分配系数以及共轭双键密切相关.通过这些参数预测药物血浆蛋白结合率具有方便快捷的优点,可以为相关药代动力学研究提供参考依据.

This study centers on automatic sleep staging with a single channel electroencephalography (EEG), with some significant findings for sleep staging. In this study, we proposed a deep learning-based network by integrating attention mechanism and bidirectional long short-term memory neural network (AT-BiLSTM) to classify wakefulness, rapid eye movement (REM) sleep and non-REM (NREM) sleep stages N1, N2 and N3. The AT-BiLSTM network outperformed five other networks and achieved an accuracy of 83.78%, a Cohen’s kappa coefficient of 0.766 and a macro F1-score of 82.14% on the PhysioNet Sleep-EDF Expanded dataset, and an accuracy of 81.72%, a Cohen’s kappa coefficient of 0.751 and a macro F1-score of 80.74% on the DREAMS Subjects dataset. The proposed AT-BiLSTM network even achieved a higher accuracy than the existing methods based on traditional feature extraction. Moreover, better performance was obtained by the AT-BiLSTM network with the frontal EEG derivations than with EEG channels located at the central, occipital or parietal lobe. As EEG signal can be easily acquired using dry electrodes on the forehead, our findings might provide a promising solution for automatic sleep scoring without feature extraction and may prove very useful for the screening of sleep disorders.

随着大数据和人工智能时代的到来,各大高校纷纷开设Python程序设计课程.该课程究竟应怎么教是每一位教师应该思考的问题.基于中国药科大学药学基地班Python课程的实践,文章提出应围绕计算思维的培养和"学以致用"的思想,展开案例驱动式教学.

Background Tai Chi (TC) mind-body exercise has been shown to reduce falls and improve balance and gait, however, few studies have evaluated the role of lower extremity muscle activation patterns in the observed benefits of TC on mobility. Purpose To perform an exploratory analysis of the association between TC training and levels of lower extremity muscle co-contraction in healthy adults during walking under single-task (ST) and cognitive dual-task (DT) conditions. Methods Surface electromyography of the anterior tibialis and lateral gastrocnemius muscles was recorded during 90 sec trials of overground ST (walking normally) and DT (walking with verbalized serial subtractions) walking. A mean co-contraction index (CCI), across all strides, was calculated based on the percentage of total muscle activity when antagonist muscles were simultaneously activated. A hybrid study design investigated long-term effects of TC via a cross-sectional comparison of 27 TC experts and 60 age-matched TC-naïve older adults. A longitudinal comparison assessed the shorter-term effects of TC; TC-naïve participants were randomly allocated to either 6 months of TC training or to usual care. Results Across all participants at baseline, greater CCI was correlated with slower gait speed under DT (β(95% CI) = -26.1(-48.6, -3.7)) but not ST (β(95% CI) = -15.4(-38.2, 7.4)) walking. Linear models adjusting for age, gender, BMI and other factors that differed at baseline indicated that TC experts exhibited lower CCI compared to TC naives under DT, but not ST conditions (ST: mean difference (95% CI) = -7.1(-15.2, 0.97); DT: mean difference (95% CI) = -10.1(-18.1, -2.4)). No differences were observed in CCI for TC-naive adults randomly assigned to 6 months of TC vs. usual care. Conclusion Lower extremity muscle co-contraction may play a role in the observed benefit of longer-term TC training on gait and postural control. Longer-duration and adequately powered randomized trials are needed to evaluate the effect of TC on neuromuscular coordination and its impact on postural control. Trial registration The randomized trial component of this study was registered at ClinicalTrials.gov ( NCT01340365 ).

Objective evaluation of sleep is crucial to recognize sleep disorders and take effective interventions accordingly. Due to the shortcomings of the conventional polysomnographic acquisition, an automatic sleep scoring method based on a single-channel electroencephalogram (EEG) is vital in sleep medicine practice. In this study, we proposed a data-driven and robust automatic sleep staging scheme based on a single-channel EEG. With the decomposition of the EEG epochs using Ensemble Empirical Mode Decomposition (EEMD), we extracted various features using statistical, time-domain and nonlinear dynamics characteristics from not only the original EEG signal but also the decomposed intrinsic mode functions (IMFs). A classification model with the eXtreme Gradient Boosting (XGBoost) algorithm was trained and tested using 5-fold cross-validation on three different databases, i.e., the Sleep-EDF database available at Physionet, the DREAMS Subjects database and the database of Sleep Heart Health Study (SHHS). The results demonstrated that the analysis of IMFs derived from EEMD could provide substantial supplement to the classification of EEG sleep stages. In the tasks of 4-class and 5-class sleep staging, the proposed method achieved an accuracy of 93.1% and 91.9% for the Sleep-EDF database, 86.4% and 83.4% for the DREAMS Subjects database, and 87.5% and 85.8% for the SHHS database, respectively. Furthermore, our observation that prefrontal EEG derivations should be optimal choices might evoke promising application of wearable EEG devices on sleep monitoring because EEG signals can be easily obtained using dry electrodes on the forehead. Additionally, the proposed method is computationally efficient and should be valuable in real-time sleep-scoring.

目的 建立基于机器学习的化合物血脑屏障通透性(logBB)预测模型,并评价其预测效果.方法 收集360个化合物的分子结构信息与相应的logBB,使用遗传算法结合极端梯度提升算法来构建化合物logBB预测模型.结果 经十折交叉验证,模型的预测平方相关系数为0.63,均方误差为0.23,表现出良好的预测性能.此外,通过对模型特征重要性的分析,总结了对化合物logBB影响程度最高的前5个分子描述符,分别是拓扑极性表面积、氢键、疏水/亲水性、油水分配系数、辛醇水分配系数.结论 模型的预测性能优于仅使用极端梯度提升算法或使用遗传算法结合支持向量机来构建的模型性能,且远优于discovery studio(DS,2016版)软件自带的预测算法性能.本研究可为治疗脑部相关疾病的药物研发提供指导性意义.

Investigation of the risk factors associated with cardiovascular disease (CVD) plays an important part in the prevention and treatment of CVD. This study investigated whether alteration in the multi-scale time irreversibility of sleeping heart rate variability (HRV) was a risk factor for cardiovascular events. The D-value, based on analysis of multi-scale increments in HRV series, was used as the measurement of time irreversibility. Eighty-four subjects from an open-access database (i.e., the Sleep Heart Health Study) were included in this study. None of them had any CVD history at baseline; 42 subjects had cardiovascular events within 1 year after baseline polysomnography and were classed as the CVD group, and the other 42 subjects in the non-CVD group were age matched with those in the CVD group and had no cardiovascular events during the 15-year follow-up period. We compared D-values of sleeping HRV between the CVD and non-CVD groups and found that the D-values of the CVD group were significantly lower than those of the non-CVD group on all 10 scales, even after adjusting for gender and body mass index. Moreover, we investigated the performance of a machine learning model to classify CVD and non-CVD subjects. The model, which was fed with a feature space based on the D-values on 10 scales and trained by a random forest algorithm, achieved an accuracy of 80.8% and a positive prediction rate of 86.7%. These results suggest that the decreased time irreversibility of sleeping HRV is an independent predictor of cardiovascular events that could be used to assist the intelligent prediction of cardiovascular events.

As a noninvasive brain function detection technique, Magnetoencephalography (MEG) has been widely used in the research of depression. By analyzing the amount of information storage, the difference of MEG information storage between patients with depression and healthy people was studied. Our analysis was carried out in the popular multiscale entropy framework, in which the time series were first "coarse-grained" on the selected time scale by low-pass filtering and down-sampling, and then its complexity was evaluated ac-cording to conditional entropy. Within this framework, we used the linear fractional integral autoregressive (ARFI) model to derive the analytical expression of information storage calculated at multiple time scales. We used the information storage expression derived from the ARFI model and then collected the information storage of MEG through positive, negative and neutral stimuli and finally calculate it. The experimental results showed that it was best to distinguish between patients with depression and healthy people through the information storage of MEG through positive stimuli, and it was best to distinguish healthy people from patients with depression at a higher frequency if it was negative or neutral stimuli.

The peak period of cardiovascular disease (CVD) is around the time of awakening in the morning, which may be related to the surge of sympathetic activity at the end of nocturnal sleep. This paper chose 140 participants as study object, 70 of which had occurred CVD events while the rest hadn't during a two-year follow-up period. A two-layer model was proposed to investigate whether hypnopompic heart rate variability (HRV) was informative to distinguish these two types of participants. In the proposed model, the extreme gradient boosting algorithm (XGBoost) was used to construct a classifier in the first layer. By evaluating the feature importance of the classifier, those features with larger importance were fed into the second layer to construct the final classifier. Three machine learning algorithms, i.e., XGBoost, random forest and support vector machine were employed and compared in the second layer to find out which one can achieve the highest performance. The results showed that, with the analysis of hypnopompic HRV, the XGBoost+XGBoost model achieved the best performance with an accuracy of 84.3%. Compared with conventional time-domain and frequency-domain features, those features derived from nonlinear dynamic analysis were more important to the model. Especially, modified permutation entropy at scale 1 and sample entropy at scale 3 were relatively important. This study might have significance for the prevention and diagnosis of CVD, as well as for the design of CVD-risk assessment system.

Insomnia is a common sleep disorder that is closely associated with the occurrence and deterioration of cardiovascular disease, depression and other diseases. The evaluation of pharmacological treatments for insomnia brings significant clinical implications. In this study, a total of 20 patients with mild insomnia and 75 healthy subjects as controls (HC) were included to explore alterations of electroencephalogram (EEG) complexity associated with insomnia and its pharmacological treatment by using multi-scale permutation entropy (MPE). All participants were recorded for two nights of polysomnography (PSG). The patients with mild insomnia received a placebo on the first night (Placebo) and temazepam on the second night (Temazepam), while the HCs had no sleep-related medication intake for either night. EEG recordings from each night were extracted and analyzed using MPE. The results showed that MPE decreased significantly from pre-lights-off to the period during sleep transition and then to the period after sleep onset, and also during the deepening of sleep stage in the HC group. Furthermore, results from the insomnia subjects showed that MPE values were significantly lower for the Temazepam night compared to MPE values for the Placebo night. Moreover, MPE values for the Temazepam night showed no correlation with age or gender. Our results indicated that EEG complexity, measured by MPE, may be utilized as an alternative approach to measure the impact of sleep medication on brain dynamics.

Surges in sympathetic activity should be a major contributor to the frequent occurrence of cardiovascular events towards the end of nocturnal sleep. We aimed to investigate whether the analysis of hypnopompic heart rate variability (HRV) could assist in the prediction of cardiovascular disease (CVD). 2217 baseline CVD-free subjects were identified and divided into CVD group and non-CVD group, according to the presence of CVD during a follow-up visit. HRV measures derived from time domain analysis, frequency domain analysis and nonlinear analysis were employed to characterize cardiac functioning. Machine learning models for both long-term and short-term CVD prediction were then constructed, based on hypnopompic HRV metrics and other typical CVD risk factors. CVD was associated with significant alterations in hypnopompic HRV. An accuracy of 81.4% was achieved in short-term prediction of CVD, demonstrating a 10.7% increase compared with long-term prediction. There was a decline of more than 6% in the predictive performance of short-term CVD outcomes without HRV metrics. The complexity of hypnopompic HRV, measured by entropy-based indices, contributed considerably to the prediction and achieved greater importance in the proposed models than conventional HRV measures. Our findings suggest that Hypnopompic HRV assists the prediction of CVD outcomes, especially the occurrence of CVD event within two years.

睡眠分期是睡眠评估的基础,在睡眠紊乱症的早期诊断和干预中起着重要的作用.本文利用集合经验模态分解对单通道脑电信号进行预处理,联合使用从分解得到的固有模态信号中提取的线性和非线性动力学等多元特性,构建了机器学习模型的输入特征空间,并最终训练出可行的睡眠自动分期模型.通过对111个健康受试者整夜睡眠数据的分期实验发现,使用本文提出的特征构建策略,能在多种经典的机器学习算法(反向传播神经网络、支持向量机、随机森林和极端梯度提升)中获得具有实用价值的睡眠自动分期模型.其中,基于极端梯度提升算法的模型在对睡眠状态进行4种分期和5种分期的任务中,准确率分别为81.0％和79.7％.

Mild cognitive impairment (MCI) is generally regarded as a prodromal stage of Alzheimer’s disease (AD). In coping with the challenges caused by AD, we analyzed resting-state functional magnetic resonance imaging data of 82 MCI subjects and 93 normal controls (NCs). The alteration of brain functional network in MCI was investigated on three scales, including global metrics, nodal characteristics, and modular properties. The results supported the existence of small worldness, hubs, and community structure in the brain functional networks of both groups. Compared with NCs, the network altered in MCI over all the three scales. In scale I, we found significantly decreased characteristic path length and increased global efficiency in MCI. Moreover, altered global network metrics were associated with cognitive level evaluated by neuropsychological assessments. In scale II, the nodal betweenness centrality of some global hubs, such as the right Crus II of cerebellar hemisphere (CERCRU2.R) and fusiform gyrus (FFG.R), changed significantly and associated with the severity and cognitive impairment in MCI. In scale III, although anatomically adjacent regions tended to be clustered into the same module regardless of group, discrepancies existed in the composition of modules in both groups, with a prominent separation of the cerebellum and a less localized organization of community structure in MCI compared with NC. Taking advantages of random forest approach, we achieved an accuracy of 91.4% to discriminate MCI patients from NCs by integrating cognitive assessments and network analysis. The importance of the used features fed into the classifier further validated the nodal characteristics of CERCRU2.R and FFG.R could be potential biomarkers in the identification of MCI. In conclusion, the present study demonstrated that the brain functional connectome data altered at the stage of MCI and could assist the automatic diagnosis of MCI patients.

ECG analysis is an important method of heart disease diagnosis. During the diagnostic process,many signal characteristics are hidden in the 12-lead ECG. To research these characteristics and improve diagnostic efficiency, it is very urgent to study the 12-lead ECG signal. In this paper, we used multifractal detrended fluctuation analysis(MFDFA) method to detect chaotic characteristics of three sets of signals, which is generated from Myocardial Infarction(MI) state, Arrhythmia state and healthy state. Calculating and analyzing the Hurst exponent, the mass exponent and the multifractal spectrum, we found that the three kinds of signals have different long-range correlation and multifractal characteristics. The result shows that the method could robustly identify patterns generated from the healthy and pathologic state. These results will assist in the intensive study of cardiac signals, guide the analysis of physiological states and provide a reference for clinical diagnosis and treatment.

Abstract Transfer entropy (TE) has been broadly used in the field of neurosciences. In this paper, the partial information decomposition algorithm is employed to decompose multivariate TE into synergistic, redundant and unique parts. In this work, the synergistic part is believed as more suitable as the computation method. We recorded the magnetoencephalogram (MEG) data of 6 subjects with depression and 13 normal subjects under different emotional stimulations, and studied the coupling between multiple symmetric channels in the frontal area in the brain of subject. The experimental results show that under different emotional stimulations, normal people present significant difference from the depression patients, especially in the right frontal area. Furthermore, under negative emotional stimulation, the difference in synergistic value between normal people and depression patients is smaller. The synergistic value of depression patients has become bigger, which indicates that the brain complexity of depression patients has grown, and their brain activities have increased.

Quantifying the complexity of the EEG signal during prolonged wakefulness and during sleep is gaining interest as an additional mean to characterize the mechanisms associated with sleep and wakefulness regulation. Here, we characterized how EEG complexity, as indexed by Multiscale Permutation Entropy (MSPE), changed progressively in the evening prior to light off and during the transition from wakefulness to sleep. We further explored whether MSPE was able to discriminate between wakefulness and sleep around sleep onset and whether MSPE changes were correlated with spectral measures of the EEG related to sleep need during concomitant wakefulness (theta power-Ptheta: 4-8 Hz). To address these questions, we took advantage of large datasets of several hundred of ambulatory EEG recordings of individual of both sexes aged 25-101 years. Results show that MSPE significantly decreases before light off (i.e. before sleep time) and in the transition from wakefulness to sleep onset. Furthermore, MSPE allows for an excellent discrimination between pre-sleep wakefulness and early sleep. Finally, we show that MSPE is correlated with concomitant Ptheta. Yet, the direction of the latter correlation changed from before light-off to the transition to sleep. Given the association between EEG complexity and consciousness, MSPE may track efficiently putative changes in consciousness preceding sleep onset. An MSPE stands as a comprehensive measure that is not limited to a given frequency band and reflects a progressive change brain state associated with sleep and wakefulness regulation. It may be an effective mean to detect when the brain is in a state close to sleep onset.