Automatic knowledge extraction from Chinese electronic medical records and rheumatoid arthritis knowledge graph construction.

Background: Knowledge graphs are a powerful tool for organizing knowledge, processing information and integrating scattered information, effectively visualizing the relationships among entities and supporting further intelligent applications. One of the critical tasks in building knowledge graphs is knowledge extraction. The existing knowledge extraction models in the Chinese medical domain usually require high-quality and large-scale manually labeled corpora for model training. In this study, we investigate rheumatoid arthritis (RA)-related Chinese electronic medical records (CEMRs) and address the automatic knowledge extraction task with a small number of annotated samples from CEMRs, from which an authoritative RA knowledge graph is constructed. Methods: After constructing the domain ontology of RA and completing manual labeling, we propose the MC-bidirectional encoder representation from transformers-bidirectional long short-term memory-conditional random field (BERT-BiLSTM-CRF) model for the named entity recognition (NER) task and the MC-BERT + feedforward neural network (FFNN) model for the entity extraction task. The pretrained language model (MC-BERT) is trained with many unlabeled medical data and fine-tuned using other medical domain datasets. We apply the established model to automatically label the remaining CEMRs, and then an RA knowledge graph is constructed based on the entities and entity relations, a preliminary assessment is conducted, and an intelligent application is presented. Results: The proposed model achieved better performance than that of other widely used models in knowledge extraction tasks, with mean F1 scores of 92.96% in entity recognition and 95.29% in relation extraction. This study preliminarily confirmed that using a pretrained medical language model could solve the problem that knowledge extraction from CEMRs requires a large number of manual annotations. An RA knowledge graph based on the above identified entities and extracted relations from 1,986 CEMRs was constructed. Experts verified the effectiveness of the constructed RA knowledge graph. Conclusions: In this paper, an RA knowledge graph based on CEMRs was established, the processes of data annotation, automatic knowledge extraction, and knowledge graph construction were described, and a preliminary assessment and an application were presented. The study demonstrated the viability of a pretrained language model combined with a deep neural network for knowledge extraction tasks from CEMRs based on a small number of manually annotated samples.

The Functional Brain Network of Subcortical and Cortical Regions Underlying Time Estimation: An Functional MRI Study.

Time estimation is fundamental for human survival. There have been increasing studies suggesting that distributed brain regions, such as the basal ganglia, cerebellum and the parietal cortex, may contribute to a dedicated neural mechanism of time estimation. However, evidence on the specific function of the subcortical and cortical brain regions and the interplay of them is scare. In this work, we explored how the subcortical and cortical networks function in time estimation during a time reproduction task using functional MRI (fMRI). Thirty healthy participants performed the time reproduction task in both auditory and visual modalities. Results showed that time estimation in visual and auditory modality recruited a subcortical-cortical brain network including the left caudate, left cerebellum, and right precuneus. Besides, the superior temporal gyrus (STG) was found essential in the difference between time estimation in visual and auditory modality. Using psychophysiological interaction (PPI) analysis, we observed an increase in the connection between left caudate and left precuneus using the left caudate as the seed region in temporal reproduction task than control task. This suggested that the left caudate is the key region connecting and transmitting information to other brain regions in the dedicated brain network of time estimation.

Comparative physiological analyses and the genetic basis reveal heat stress responses mechanism among different Betula luminifera populations.

Betula luminifera is a subtropical fast-growing timber species with high economic value. However, along with global warming, heat stress become one of the main environmental variables that limit the productivity of B. luminifera, and the response of diverse geographic populations to high temperatures is still unclear. In order to offer a comprehensive understanding of the behavior of B. luminifera under heat stress, the physiological responses of six B. luminifera populations (across the core distribution area) were described in this work in an integrated viewpoint. The results showed that a multi-level physiological regulatory network may exist in B. luminifera, the first response was the activity of resistant enzymes [e.g., peroxidase (POD)] at a preliminary stage of 2 h heat stress, and then the proline (osmoregulation substance) content began to increase after 24 h of continuous high-temperature treatment. In addition, photosynthesis was stronlgly affected by heat stress, and the net photosynthetic rate (Pn ) showed a downward trend under heat treatment in all six B. luminifera populations. Interestingly, although the physiological change patterns of the six B. luminifera populations were relatively consistent for the same parameter, there were obvious differences among different populations. Comprehensive analysis revealed that the physiological response of Rongshui (RS) was the most stable, and this was the representative B. luminifera population. Illumina RNA-seq analysis was applied to reveal the specific biological process of B. luminifera under heat stress using the RS population, and a total of 116,484 unigenes were obtained. The differentially expressed genes (DEGs) between different time periods under heat stress were enriched in 34 KEGG pathways, and the limonene and pinene degradation pathway was commonly enriched in all pairwise comparisons. Moreover, transcription factors including bHLH (basic helix-loop-helix), MYB, WRKY, and NAC (NAM, ATAF1/2, and CUC2) were identified. In this study, the physiological response and tolerance mechanisms of B. luminifera under high temperature stress were revealed, which can conducive to the basis of B. luminifera selection and resistance assessment for cultivation and breeding.

Unpredictable fearful stimuli disrupt timing activities: Evidence from event-related potentials.

The present study investigated the effect of an imminent fearful stimulus on an ongoing temporal task. Participants judged the duration of a blank temporal interval followed by a fearful or a neutral image. Results showed an underestimation of the duration in the fearful condition relative to the neutral condition, but only when the occurrence of the fearful image was difficult to predict. ERPs results for the blank temporal interval found no effect of the fearful stimulus on the contingent negative variation (CNV) amplitude in the clock stage. However, after the image onset, there was a larger P1 for the fearful relative to the neutral condition. Although this effect was indistinguishable regardless of whether the fearful event could be easily predicted, a late positive potential (LPP) component displayed larger amplitude only for unpredictable fearful stimuli. The time-frequency results showed enhanced delta-theta power (0.5-7.5 Hz) for the unpredictable fearful stimuli in the late stage. Importantly, the enhanced delta-theta rhythm correlated negatively with the duration judgments. Together, these results suggest that an unpredictable fearful event might divert more attention away from the counting process in the working memory stage, resulting in missing ticks and temporal underestimation.

Reexamining the neural network involved in perception of facial expression: A meta-analysis.

Perception of facial expression is essential for social interactions. Although a few competing models have enjoyed some success to map brain regions, they are also facing difficult challenges. The current study used an updated activation likelihood estimation (ALE) method of meta-analysis to explore the involvement of brain regions in facial expression processing. The sample contained 96 functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) studies of healthy adults with the results of whole-brain analyses. The key findings revealed that the ventral pathway, especially the left fusiform face area (FFA) region, was more responsive to facial expression. The left posterior FFA showed strong involvement when participants passively viewing emotional faces without being asked to judge the type of expression or other attributes of the stimuli. Through meta-analytic connectivity modeling (MACM) of the main brain regions in the ventral pathway, we constructed a co-activating neural network as a revised model of facial expression processing that assigns prominent roles to the amygdala, FFA, the occipital gyrus, and the inferior frontal gyrus.

The Preponderant Role of Fusiform Face Area for the Facial Expression Confusion Effect: An MEG Study.

Although the recognition of facial expressions seems automatic and effortless, discrimination of expressions can still be error prone. Common errors are often due to visual similarities between some expressions (e.g., fear and surprise). However, little is known about the neural mechanisms underlying such a confusion effect. To address this question, we recorded the magnetoencephalography (MEG) while participants judged facial expressions that were either easily confused with or easily distinguished from other expressions. The results showed that the fusiform face area (FFA), rather than the posterior superior temporal sulcus (pSTS), played a preponderant role in discriminating confusable facial expressions. No difference between high confusion and low confusion conditions was observed on the M170 component in either the FFA or the pSTS, whilst a difference between two conditions started to emerge in the late positive potential (LPP), with the low confusion condition eliciting a larger LPP amplitude in the FFA. In addition, the power of delta was stronger in the time window of LPP component. This confusion effect was reflected in the FFA, which might be associated with the perceptual-to-conceptual shift.