This paper is dedicated to the design and evaluation of the first AMR parser tailored for clinical notes. Our objective was to facilitate the precise transformation of the clinical notes into structured AMR expressions, thereby enhancing the interpretability and usability of clinical text data at scale. Leveraging the colon cancer dataset from the Temporal Histories of Your Medical Events (THYME) corpus, we adapted a state-of-the-art AMR parser utilizing continuous training. Our approach incorporates data augmentation techniques to enhance the accuracy of AMR structure predictions. Notably, through this learning strategy, our parser achieved an impressive F1 score of 88% on the THYME corpus’s colon cancer dataset. Moreover, our research delved into the efficacy of data required for domain adaptation within the realm of clinical notes, presenting domain adaptation data requirements for AMR parsing. This exploration not only underscores the parser’s robust performance but also highlights its potential in facilitating a deeper understanding of clinical narratives through structured semantic representations.

The 2024 Shared Task on Chemotherapy Treatment Timeline Extraction aims to advance the state of the art of clinical event timeline extraction from the Electronic Health Records (EHRs). Specifically, this edition focuses on chemotherapy event timelines from EHRs of patients with breast, ovarian and skin cancers. These patient-level timelines present a novel challenge which involves tasks such as the extraction of relevant events, time expressions and temporal relations from each document and then summarizing over the documents. De-identified EHRs for 57,530 patients with breast and ovarian cancer spanning 2004-2020, and approximately 15,946 patients with melanoma spanning 2010-2020 were made available to participants after executing a Data Use Agreement. A subset of patients is annotated for gold entities, time expressions, temporal relations and patient-level timelines. The rest is considered unlabeled data. In Subtask1, gold chemotherapy event mentions and time expressions are provided (along with the EHR notes). Participants are asked to build the patient-level timelines using gold annotations as input. Thus, the subtask seeks to explore the topics of temporal relations extraction and timeline creation if event and time expression input is perfect. In Subtask2, which is the realistic real-world setting, only EHR notes are provided. Thus, the subtask aims at developing an end-to-end system for chemotherapy treatment timeline extraction from patient’s EHR notes. There were 18 submissions for Subtask 1 and 9 submissions for Subtask 2. The organizers provided a baseline system. The teams employed a variety of methods including Logistic Regression, TF-IDF, n-grams, transformer models, zero-shot prompting with Large Language Models (LLMs), and instruction tuning. The gap in performance between prompting LLMs and finetuning smaller-sized LMs indicates that for a challenging task such as patient-level chemotherapy timeline extraction, more sophisticated LLMs or prompting techniques are necessary in order to achieve optimal results as finetuing smaller-sized LMs outperforms by a wide margin.

The bias-variance tradeoff is the idea that learning methods need to balance model complexity with data size to minimize both under-fitting and over-fitting. Recent empirical work and theoretical analysis with over-parameterized neural networks challenges the classic bias-variance trade-off notion suggesting that no such trade-off holds: as the width of the network grows, bias monotonically decreases while variance initially increases followed by a decrease. In this work, we first provide a variance decomposition-based justification criteria to examine whether large pretrained neural models in a fine-tuning setting are generalizable enough to have low bias and variance. We then perform theoretical and empirical analysis using ensemble methods explicitly designed to decrease variance due to optimization. This results in essentially a two-stage fine-tuning algorithm that first ratchets down bias and variance iteratively, and then uses a selected fixed-bias model to further reduce variance due to optimization by ensembling. We also analyze the nature of variance change with the ensemble size in low- and high-resource classes. Empirical results show that this two-stage method obtains strong results on SuperGLUE tasks and clinical information extraction tasks. Code and settings are available: https://github.com/christa60/bias-var-fine-tuning-plms.git

Understanding temporal relationships in text from electronic health records can be valuable for many important downstream clinical applications. Since Clinical TempEval 2017, there has been little work on end-to-end systems for temporal relation extraction, with most work focused on the setting where gold standard events and time expressions are given. In this work, we make use of a novel multi-headed attention mechanism on top of a pre-trained transformer encoder to allow the learning process to attend to multiple aspects of the contextualized embeddings. Our system achieves state of the art results on the THYME corpus by a wide margin, in both the in-domain and cross-domain settings.

In-context learning (ICL) is a new learning paradigm that has gained popularity along with the development of large language models. In this work, we adapt a recently proposed hardness metric, pointwise 𝒱-usable information (PVI), to an in-context version (in-context PVI). Compared to the original PVI, in-context PVI is more efficient in that it requires only a few exemplars and does not require fine-tuning. We conducted a comprehensive empirical analysis to evaluate the reliability of in-context PVI. Our findings indicate that in-context PVI estimates exhibit similar characteristics to the original PVI. Specific to the in-context setting, we show that in-context PVI estimates remain consistent across different exemplar selections and numbers of shots. The variance of in-context PVI estimates across different exemplar selections is insignificant, which suggests that in-context PVI estimates are stable. Furthermore, we demonstrate how in-context PVI can be employed to identify challenging instances. Our work highlights the potential of in-context PVI and provides new insights into the capabilities of ICL.

We explore temporal dependency graph (TDG) parsing in the clinical domain. We leverage existing annotations on the THYME dataset to semi-automatically construct a TDG corpus. Then we propose a new natural language inference (NLI) approach to TDG parsing, and evaluate it both on general domain TDGs from wikinews and the newly constructed clinical TDG corpus. We achieve competitive performance on general domain TDGs with a much simpler model than prior work. On the clinical TDGs, our method establishes the first result of TDG parsing on clinical data with 0.79/0.88 micro/macro F1.

In this paper, we investigate ensemble methods for fine-tuning transformer-based pretrained models for clinical natural language processing tasks, specifically temporal relation extraction from the clinical narrative. Our experimental results on the THYME data show that ensembling as a fine-tuning strategy can further boost model performance over single learners optimized for hyperparameters. Dynamic snapshot ensembling is particularly beneficial as it fine-tunes a wide array of parameters and results in a 2.8% absolute improvement in F1 over the base single learner.

Sequence-to-sequence models are appealing because they allow both encoder and decoder to be shared across many tasks by formulating those tasks as text-to-text problems. Despite recently reported successes of such models, we find that engineering input/output representations for such text-to-text models is challenging. On the Clinical TempEval 2016 relation extraction task, the most natural choice of output representations, where relations are spelled out in simple predicate logic statements, did not lead to good performance. We explore a variety of input/output representations, with the most successful prompting one event at a time, and achieving results competitive with standard pairwise temporal relation extraction systems.

Transformer-based neural language models have led to breakthroughs for a variety of natural language processing (NLP) tasks. However, most models are pretrained on general domain data. We propose a methodology to produce a model focused on the clinical domain: continued pretraining of a model with a broad representation of biomedical terminology (PubMedBERT) on a clinical corpus along with a novel entity-centric masking strategy to infuse domain knowledge in the learning process. We show that such a model achieves superior results on clinical extraction tasks by comparing our entity-centric masking strategy with classic random masking on three clinical NLP tasks: cross-domain negation detection, document time relation (DocTimeRel) classification, and temporal relation extraction. We also evaluate our models on the PubMedQA dataset to measure the models’ performance on a non-entity-centric task in the biomedical domain. The language addressed in this work is English.

We present work on extraction of radiotherapy treatment information from the clinical narrative in the electronic medical records. Radiotherapy is a central component of the treatment of most solid cancers. Its details are described in non-standardized fashions using jargon not found in other medical specialties, complicating the already difficult task of manual data extraction. We examine the performance of several state-of-the-art neural methods for relation extraction of radiotherapy treatment details, with a goal of automating detailed information extraction. The neural systems perform at 0.82-0.88 macro-average F1, which approximates or in some cases exceeds the inter-annotator agreement. To the best of our knowledge, this is the first effort to develop models for radiotherapy relation extraction and one of the few efforts for relation extraction to describe cancer treatment in general.

We present refinements over existing temporal relation annotations in the Electronic Medical Record clinical narrative. We refined the THYME corpus annotations to more faithfully represent nuanced temporality and nuanced temporal-coreferential relations. The main contributions are in re-defining CONTAINS and OVERLAP relations into CONTAINS, CONTAINS-SUBEVENT, OVERLAP and NOTED-ON. We demonstrate that these refinements lead to substantial gains in learnability for state-of-the-art transformer models as compared to previously reported results on the original THYME corpus. We thus establish a baseline for the automatic extraction of these refined temporal relations. Although our study is done on clinical narrative, we believe it addresses far-reaching challenges that are corpus- and domain- agnostic.

Recently BERT has achieved a state-of-the-art performance in temporal relation extraction from clinical Electronic Medical Records text. However, the current approach is inefficient as it requires multiple passes through each input sequence. We extend a recently-proposed one-pass model for relation classification to a one-pass model for relation extraction. We augment this framework by introducing global embeddings to help with long-distance relation inference, and by multi-task learning to increase model performance and generalizability. Our proposed model produces results on par with the state-of-the-art in temporal relation extraction on the THYME corpus and is much “greener” in computational cost.

This paper discusses a cross-document coreference annotation schema that was developed to further automatic extraction of timelines in the clinical domain. Lexical senses and coreference choices are determined largely by context, but cross-document work requires reasoning across contexts that are not necessarily coherent. We found that an annotation approach that relies less on context-guided annotator intuitions and more on schematic rules was most effective in creating meaningful and consistent cross-document relations.

Classic methods for clinical temporal relation extraction focus on relational candidates within a sentence. On the other hand, break-through Bidirectional Encoder Representations from Transformers (BERT) are trained on large quantities of arbitrary spans of contiguous text instead of sentences. In this study, we aim to build a sentence-agnostic framework for the task of CONTAINS temporal relation extraction. We establish a new state-of-the-art result for the task, 0.684F for in-domain (0.055-point improvement) and 0.565F for cross-domain (0.018-point improvement), by fine-tuning BERT and pre-training domain-specific BERT models on sentence-agnostic temporal relation instances with WordPiece-compatible encodings, and augmenting the labeled data with automatically generated “silver” instances.

Automatic identification of spurious instances (those with potentially wrong labels in datasets) can improve the quality of existing language resources, especially when annotations are obtained through crowdsourcing or automatically generated based on coded rankings. In this paper, we present effective approaches inspired by queueing theory and psychology of learning to automatically identify spurious instances in datasets. Our approaches discriminate instances based on their “difficulty to learn,” determined by a downstream learner. Our methods can be applied to any dataset assuming the existence of a neural network model for the target task of the dataset. Our best approach outperforms competing state-of-the-art baselines and has a MAP of 0.85 and 0.22 in identifying spurious instances in synthetic and carefully-crowdsourced real-world datasets respectively.

Neural network models are oftentimes restricted by limited labeled instances and resort to advanced architectures and features for cutting edge performance. We propose to build a recurrent neural network with multiple semantically heterogeneous embeddings within a self-training framework. Our framework makes use of labeled, unlabeled, and social media data, operates on basic features, and is scalable and generalizable. With this method, we establish the state-of-the-art result for both in- and cross-domain for a clinical temporal relation extraction task.

Clinical TempEval 2017 aimed to answer the question: how well do systems trained on annotated timelines for one medical condition (colon cancer) perform in predicting timelines on another medical condition (brain cancer)? Nine sub-tasks were included, covering problems in time expression identification, event expression identification and temporal relation identification. Participant systems were evaluated on clinical and pathology notes from Mayo Clinic cancer patients, annotated with an extension of TimeML for the clinical domain. 11 teams participated in the tasks, with the best systems achieving F1 scores above 0.55 for time expressions, above 0.70 for event expressions, and above 0.40 for temporal relations. Most tasks observed about a 20 point drop over Clinical TempEval 2016, where systems were trained and evaluated on the same domain (colon cancer).

We present a novel approach for training artificial neural networks. Our approach is inspired by broad evidence in psychology that shows human learners can learn efficiently and effectively by increasing intervals of time between subsequent reviews of previously learned materials (spaced repetition). We investigate the analogy between training neural models and findings in psychology about human memory model and develop an efficient and effective algorithm to train neural models. The core part of our algorithm is a cognitively-motivated scheduler according to which training instances and their “reviews” are spaced over time. Our algorithm uses only 34-50% of data per epoch, is 2.9-4.8 times faster than standard training, and outperforms competing state-of-the-art baselines. Our code is available at scholar.harvard.edu/hadi/RbF/.

We experiment with neural architectures for temporal relation extraction and establish a new state-of-the-art for several scenarios. We find that neural models with only tokens as input outperform state-of-the-art hand-engineered feature-based models, that convolutional neural networks outperform LSTM models, and that encoding relation arguments with XML tags outperforms a traditional position-based encoding.

Detecting negated concepts in clinical texts is an important part of NLP information extraction systems. However, generalizability of negation systems is lacking, as cross-domain experiments suffer dramatic performance losses. We examine the performance of multiple unsupervised domain adaptation algorithms on clinical negation detection, finding only modest gains that fall well short of in-domain performance.

Token sequences are often used as the input for Convolutional Neural Networks (CNNs) in natural language processing. However, they might not be an ideal representation for time expressions, which are long, highly varied, and semantically complex. We describe a method for representing time expressions with single pseudo-tokens for CNNs. With this method, we establish a new state-of-the-art result for a clinical temporal relation extraction task.

This article discusses the requirements of a formal specification for the annotation of temporal information in clinical narratives. We discuss the implementation and extension of ISO-TimeML for annotating a corpus of clinical notes, known as the THYME corpus. To reflect the information task and the heavily inference-based reasoning demands in the domain, a new annotation guideline has been developed, “the THYME Guidelines to ISO-TimeML (THYME-TimeML)”. To clarify what relations merit annotation, we distinguish between linguistically-derived and inferentially-derived temporal orderings in the text. We also apply a top performing TempEval 2013 system against this new resource to measure the difficulty of adapting systems to the clinical domain. The corpus is available to the community and has been proposed for use in a SemEval 2015 task.

This paper presents the evaluation of the dictionary look-up component of Mayo Clinic’s Information Extraction system. The component was tested on a corpus of 160 free-text clinical notes which were manually annotated with the named entity disease. This kind of clinical text presents many language challenges such as fragmented sentences and heavy use of abbreviations and acronyms. The dictionary used for this evaluation was a subset of SNOMED-CT with semantic types corresponding to diseases/disorders without any augmentation. The algorithm achieves an F-score of 0.56 for exact matches and F-scores of 0.76 and 0.62 for right and left-partial matches respectively. Machine learning techniques are currently under investigation to improve this task.

We report on the construction of a gold-standard dataset consisting of annotated clinical notes suitable for evaluating our biomedical named entity recognition system. The dataset is the result of consensus between four human annotators and contains 1,556 annotations on 160 clinical notes using 658 unique concept codes from SNOMED-CT corresponding to human disorders. Inter-annotator agreement was calculated on annotations from 100 of the documents for span (90.9%), concept code (81.7%), context (84.8%), and status (86.0%) agreement. Complete agreement for span, concept code, context, and status was 74.6%. We found that creating a consensus set based on annotations from two independently-created annotation sets can reduce inter-annotator disagreement by 32.3%. We found little benefit to pre-annotating the corpus with a third-party named entity recognizer.