Automated Preamble Detection in Dictated Medical Reports

Salloum, Wael; Finley, Greg; Edwards, Erik; Miller, Mark A.; Suendermann-Oeft, David

doi:10.18653/v1/w17-2336

Cited by 3 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We measure PER of the MTPP against a baseline system, which was also developed internally within EMR.AI for specific use with clinical dictations. The baseline system employs a modular pipeline, where each module is responsible for a particular transformation-for instance, one detects a metadata-heavy "preamble" in the dictation (Salloum et al, 2017a); another converts spelledout numbers to numerals, dates, etc. Some components of the system are rule based, while others rely on machine learning.…”

Section: Discussionmentioning

confidence: 99%

From dictations to clinical reports using machine translation

Finley

Salloum

Sadoughi³

et al. 2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

Self Cite

View full text Add to dashboard Cite

A typical workflow to document clinical encounters entails dictating a summary, running speech recognition, and post-processing the resulting text into a formatted letter. Post-processing entails a host of transformations including punctuation restoration, truecasing, marking sections and headers, converting dates and numerical expressions, parsing lists, etc. In conventional implementations, most of these tasks are accomplished by individual modules. We introduce a novel holistic approach to post-processing that relies on machine callytranslation. We show how this technique outperforms an alternative conventional system-even learning to correct speech recognition errors during post-processingwhile being much simpler to maintain.

show abstract

Section: Discussionmentioning

confidence: 99%

From dictations to clinical reports using machine translation

Finley

Salloum

Sadoughi³

et al. 2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

Self Cite

View full text Add to dashboard Cite

show abstract

“…For example, Sadoughi et al used unidirectional long-short term memory (LSTM) units. And Salloum et al proposed using bi-direction LSTM to detect sections while converting from medical dictations into clinical reports [16,17]. In a recent study, Rosenthal et al applied recurrent neural network (RNN) or the fine-tune BERT model using gated recurrence units trained with medical literature.…”

Section: Approaches For Section Detectionmentioning

confidence: 99%

Assessing document section heterogeneity across multiple electronic health record systems for computational phenotyping: A case study of heart-failure phenotyping algorithm

et al. 2023

View full text Add to dashboard Cite

Background The incorporation of information from clinical narratives is critical for computational phenotyping. The accurate interpretation of clinical terms highly depends on their associated context, especially the corresponding clinical section information. However, the heterogeneity across different Electronic Health Record (EHR) systems poses challenges in utilizing the section information. Objectives Leveraging the eMERGE heart failure (HF) phenotyping algorithm, we assessed the heterogeneity quantitatively through the performance comparison of machine learning (ML) classifiers which map clinical sections containing HF-relevant terms across different EHR systems to standard sections in Health Level 7 (HL7) Clinical Document Architecture (CDA). Methods We experimented with both random forest models with sentence-embedding features and bidirectional encoder representations from transformers models. We trained MLs using an automated labeled corpus from an EHR system that adopted HL7 CDA standard. We assessed the performance using a blind test set (n = 300) from the same EHR system and a gold standard (n = 900) manually annotated from three other EHR systems. Results The F-measure of those ML models varied widely (0.00–0.91%), indicating MLs with one tuning parameter set were insufficient to capture sections across different EHR systems. The error analysis indicates that the section does not always comply with the corresponding standardized sections, leading to low performance. Conclusions We presented the potential use of ML techniques to map the sections containing HF-relevant terms in multiple EHR systems to standard sections. However, the findings suggested that the quality and heterogeneity of section structure across different EHRs affect applications due to the poor adoption of documentation standards.

show abstract

“…As the predicted probability score falls between 0 and 1, the tokens predicted with high Is-intro probabilities may be found throughout a document (see Figure 3). We therefore create a simple maximum difference algorithm inspired by Salloum et al [22] to identify the best segmentation boundaries . We evaluate how likely each token is the introduction's start position by averaging the scores of tokens before and after it:…”

Section: Approachmentioning

confidence: 99%

“…Inspired by recent works on text segmentation using neural network models [3,14,22], we formulate our task as a supervised sequence labeling task. We train our models to label each token in the text, and find the best split position based on the token labels, using fine-tuning over a pre-trained BERT model [10].…”

Section: Introductionmentioning

confidence: 99%

Identifying Introductions in Podcast Episodes from Automatically Generated Transcripts

Jing¹,

Schneck²,

Egan³

et al. 2021

Preprint

View full text Add to dashboard Cite

As the volume of long-form spoken-word content such as podcasts explodes, many platforms desire to present short, meaningful, and logically coherent segments extracted from the full content. Such segments can be consumed by users to sample content before diving in, as well as used by the platform to promote and recommend content. However, little published work is focused on the segmentation of spoken-word content, where the errors (noise) in transcripts generated by automatic speech recognition (ASR) services poses many challenges. Here we build a novel dataset of complete transcriptions of over 400 podcast episodes, in which we label the position of introductions in each episode. These introductions contain information about the episodes' topics, hosts, and guests, providing a valuable summary of the episode content, as it is created by the authors. We further augment our dataset with word substitutions to increase the amount of available training data. We train three Transformer models based on the pre-trained BERT [10] and different augmentation strategies, which achieve significantly better performance compared with a static embedding model, showing that it is possible to capture generalized, larger-scale structural information from noisy, loosely-organized speech data. This is further demonstrated through an analysis of the models' inner architecture. Our methods and dataset can be used to facilitate future work on the structure-based segmentation of spoken-word content.

show abstract

Automated Preamble Detection in Dictated Medical Reports

Cited by 3 publications

References 27 publications

From dictations to clinical reports using machine translation

From dictations to clinical reports using machine translation

Assessing document section heterogeneity across multiple electronic health record systems for computational phenotyping: A case study of heart-failure phenotyping algorithm

Identifying Introductions in Podcast Episodes from Automatically Generated Transcripts

Contact Info

Product

Resources

About