Search for: All records

Abstract Tissue engineered cardiac patches have great potential as a regenerative therapy for myocardial infarction. Yet, the mutual interaction of cardiac patches with healthy tissue has not been completely understood. Here, we investigated the impact of acellular and cellular patches on a beating two-dimensional (2D) cardiac cell layer, and the effect of the beating of this layer on the cells encapsulated in the patch. We cultured human-induced pluripotent stem cell-derived cardiomyocytes (iCMs) on a coverslip and placed gelatin methacryloyl hydrogel alone or with encapsulated iCMs to create acellular and cellular patches, respectively. When the acellular patch was placed on the cardiac cell layer, the beating characteristics and Ca+2 handling properties reduced, whereas placing the cellular patch restored these characteristics. To better understand the effects of the cyclic contraction and relaxation induced by the beating cardiac cell layer on the patch placed on top of it, a simulation model was developed, and the calculated strain values were in agreement with the values measured experimentally. Moreover, this dynamic culture induced by the beating 2D iCM layer on the iCMs encapsulated in the cellular patch improved their beating velocity and frequency. Additionally, the encapsulated iCMs were observed to be coupled with the underlying beating 2D iCM layer. Overall, this study provides a detailed investigation on the mutual relationship of acellular/cellular patches with the beating 2D iCM layer, understanding of which would be valuable for developing more advanced cardiac patches. 

We propose a novel task, G4C (Goal-driven Guidance Generation in Grounded Communication), for studying goal-driven and grounded natural language interactions. Specifically, we choose Dungeons and Dragons (D&D) -- a role-playing game consisting of multiple player characters and a Dungeon Master (DM) who collaborate to achieve a set of goals that are beneficial to the players -- as a testbed for this task. Here, each of the player characters is a student, with their own personas and abilities, and the DM is the teacher, an arbitrator of the rules of the world and responsible for assisting and guiding the students towards a global goal. We propose a theory-of-mind-inspired methodology for training such a DM with reinforcement learning (RL), where a DM: (1) learns to predict how the players will react to its utterances using a dataset of D&D dialogue transcripts; and (2) uses this prediction as a reward function providing feedback on how effective these utterances are at guiding the players towards a goal. Human and automated evaluations show that a DM trained with RL to generate guidance by incorporating a theory-of-mind of the players significantly improves the players' ability to achieve goals grounded in their shared world. 

What conditions enable novel intellectual contributions to diffuse and become integrated into later scientific work? Prior work tends to focus on whole cultural products, such as patents and articles, and emphasizes external social factors as important. This article focuses on concepts as reflections of ideas, and we identify the combined influence that social factors and internal intellectual structures have on ideational diffusion. To develop this perspective, we use computational techniques to identify nearly 60,000 new ideas introduced over two decades (1993 to 2016) in the Web of Science and follow their diffusion across 38 million later publications. We find new ideas diffuse more widely when they socially and intellectually resonate. New ideas become core concepts of science when they reach expansive networks of unrelated authors, achieve consistent intellectual usage, are associated with other prominent ideas, and fit with extant research traditions. These ecological conditions play an increasingly decisive role later in an idea’s career, after their relations with the environment are established. This work advances the systematic study of scientific ideas by moving beyond products to focus on the content of ideas themselves and applies a relational perspective that takes seriously the contingency of their success. 

Cardiomyocytes (CMs) and fibroblast cells are two essential elements for cardiac tissue structure and function. The interactions between them can alter cardiac electrophysiology and thus contribute to cardiac diseases, such as arrhythmogenesis. One possible explanation is that fibroblasts can directly affect cardiac electrophysiology through electrical coupling with CMs. Therefore, detecting the electrical activities in the CM-fibroblast network is vital for understanding the coupling dynamics among them. Current commercialized platforms for studying cardiac electrophysiology utilize planar microelectrode arrays (MEAs) to record the extracellular field potential (FP) in real-time, but the prearranged electrode configuration highly limits the measurement capabilities at specific locations. Here, we report a custom-designed MEA device with a novel micropatterning method to construct a controlled network of neonatal rat CMs (rCMs) and fibroblast connections for monitoring the electrical activity of rCM-fibroblast co-cultures in a spatially controlled fashion. For the micropatterning of the co-culture, surface topographical features and mobile blockers were used to control the initial attachment locations of a mixture of rCMs and fibroblasts, to form separate beating rCM-fibroblast clusters while leaving empty space for fibroblast growth to connect these clusters. Once the blockers are removed, the proliferating fibroblasts connect and couple the separate beating clusters. Using this method, electrical activity of both rCMs and human-induced-pluripotent-stem-cell-derived cardiomyocytes (iCMs) was examined. The coupling dynamics were studied through the extracellular FP and impedance profile recorded from the MEA device, indicating that the fibroblast bridge provided an RC-type coupling of physically separate rCM-containing clusters and enabled synchronization of these clusters. 

Increasing concerns and regulations about data privacy and sparsity necessitate the study of privacy-preserving, decentralized learning methods for natural language processing (NLP) tasks. Federated learning (FL) provides promising approaches for a large number of clients (e.g., personal devices or organizations) to collaboratively learn a shared global model to benefit all clients while allowing users to keep their data locally. Despite interest in studying FL methods for NLP tasks, a systematic comparison and analysis is lacking in the literature. Herein, we present the FedNLP, a benchmarking framework for evaluating federated learning methods on four different task formulations: text classification, sequence tagging, question answering, and seq2seq. We propose a universal interface between Transformer-based language models (e.g., BERT, BART) and FL methods (e.g., FedAvg, FedOPT, etc.) under various non-IID partitioning strategies. Our extensive experiments with FedNLP provide empirical comparisons between FL methods and helps us better understand the inherent challenges of this direction. The comprehensive analysis points to intriguing and exciting future research aimed at developing FL methods for NLP tasks. 

Hierarchical multi-label text classification (HMTC) aims to tag each document with a set of classes from a class hierarchy. Most existing HMTC methods train classifiers using massive human-labeled documents, which are often too costly to obtain in real-world applications. In this paper, we explore to conduct HMTC based on only class surface names as supervision signals. We observe that to perform HMTC, human experts typically first pinpoint a few most essential classes for the document as its “core classes”, and then check core classes’ ancestor classes to ensure the coverage. To mimic human experts, we propose a novel HMTC framework, named TaxoClass. Specifically, TaxoClass (1) calculates document-class similarities using a textual entailment model, (2) identifies a document’s core classes and utilizes confident core classes to train a taxonomyenhanced classifier, and (3) generalizes the classifier via multi-label self-training. Our experiments on two challenging datasets show TaxoClass can achieve around 0.71 Example- F1 using only class names, outperforming the best previous method by 25%. 

Entity set expansion and synonym discovery are two critical NLP tasks. Previous studies accomplish them separately, without exploring their interdependences. In this work, we hypothesize that these two tasks are tightly coupled because two synonymous entities tend to have similar likelihoods of belonging to various semantic classes. This motivates us to design SynSetExpan, a novel framework that enables two tasks to mutually enhance each other. SynSetExpan uses a synonym discovery model to include popular entities’ infrequent synonyms into the set, which boosts the set expansion recall. Meanwhile, the set expansion model, being able to determine whether an entity belongs to a semantic class, can generate pseudo training data to fine-tune the synonym discovery model towards better accuracy. To facilitate the research on studying the interplays of these two tasks, we create the first large-scale Synonym-Enhanced Set Expansion (SE2) dataset via crowdsourcing. Extensive experiments on the SE2 dataset and previous benchmarks demonstrate the effectiveness of SynSetExpan for both entity set expansion and synonym discovery tasks.