Natural language processing of electronic health records is superior to billing codes to identify symptom burden in hemodialysis patients

Chan, Lili; Beers, Kelly; Yau, Amy A.; Chauhan, Kinsuk; Duffy, Áine; Chaudhary, Kumardeep; Debnath, Neha; Saha, Aparna; Pattharanitima, Pattharawin; Cho, Judy H.; Kotanko, Peter; Federman, Alex D.; Coca, Steven G.; Vleck, Tielman Van; Nadkarni, Girish N.

doi:10.1016/j.kint.2019.10.023

Cited by 32 publications

(26 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the utility of NLP demonstrated in other disciplines, we should consider whether NLP might be leveraged more broadly in the field of nephrology. In the accompanying article, Chan et al 5 When exploring the applicability of NLP to a clinical scenario, there are multiple issues that may impact system performance, which we highlight below. An initial important consideration is the clinical question(s) being addressed, which will define how the NLP output will be used and determine the NLP system requirements.…”

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

“…The accompanying study presents an appropriate use case that leverages NLP to efficiently identify symptomatic hemodialysis patients at a population level. Chan et al 5 report that NLP is more sensitive than administrative codes for detecting all 7 symptoms, but less specific for a subset of symptoms. In this scenario, attributing a symptom to a patient if it is not actually present (i.e., a false positive) is unlikely to be of major clinical consequence.…”

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

“…With careful design and optimization, NLP systems have achieved specificities exceeding 0.90 for some clinical use cases. 3 While NLP may entail a wide variety of tasks ranging from understanding to generating language (e.g., Amazon Alexa, Google Voice, Apple Siri), Chan et al 5 used NLP for a task known as "concept" or "entity" recognition. The goal of this task is to identify mentions of concepts of interest within a note.…”

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

“…NLP systems can address this challenge by referencing standardized collections of clinical concepts, their derivatives, and potential phrase variants such as Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT). The accompanying article provides an example using SNOMED-CT to identify variants of the "cramp" symptom c o m m e n t a r y (see Supplemental Figure S6 in Chan et al 5 ).…”

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

See 3 more Smart Citations

Considerations for advancing nephrology research and practice through natural language processing

Parr

Gobbel

2020

Kidney International

View full text Add to dashboard Cite

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

Section: Considerations For Advancing Nephrology Research and Practicmentioning

confidence: 99%

See 2 more Smart Citations

Considerations for advancing nephrology research and practice through natural language processing

Parr

Gobbel

2020

Kidney International

View full text Add to dashboard Cite

“…Natural Language Processing (NLP) is gaining relevance within the clinical documentation services to cope with extensive information conveyed by Electronic Health Records (EHRs). Healthcare data is getting increasingly larger and complex to process [1], but evidence shows its usefulness in such different sectors as Adverse Drug Reaction extraction [2,3] and identification of complex symptoms, assessed in several cohorts of patients in hemodialysis [4], as well as relevant symptoms in patients with schizophrenia [5], and breast cancer [6], and the creation of phenotypes to characterise patients [7,8].…”

Section: Introductionmentioning

confidence: 99%

Extreme Multi-Label ICD Classification: Sensitivity to Hospital Service and Time

2020

View full text Add to dashboard Cite

This work deals with clinical text mining for automatic classification of Electronic Health Records (EHRs) with respect to the International Classification of Diseases (ICD). ICD is the international standard for the identification of diseases and health conditions in EHRs and the foundation for reporting health statistics. Machine learning-based techniques have proven robust to infer classification models from EHRs. Since each EHR tends to involve multiple diseases, multi-label classification is required. The concern in this work is the versatility of the models inferred and their ability to generalise in two ways: as time goes ahead and across hospital services or health specialties. Indeed, in this work, we show the capabilities of a Bidirectional Recurrent Neural Network (RNN) with GRU units and ELMo embeddings on two corpora (a corpus comprising a set of EHRs within the Basque Health System, namely Osakidetza, and the well-known MIMIC-III corpus). To delve into and assess the versatility of the models, we focus on their resilience across hospital admissions taken over two different years and also across six distinct hospital services. In addition, we paid attention to the classification performance to estimate ICD codes of different granularity (e.g. with or without essential modifiers). Our best results are 39.55% and 47.28% F-Score for the Osakidetza and MIMIC-III datasets respectively, with the original main label-sets. Regarding the models evaluated per specialty, the most remarkable results are 57.00% and 72.74% F-Score, in the Cardiology and Nephrology medical services respectively. INDEX TERMS Extreme multi-label classification, Electronic Health Records, International Classification of Diseases, Classification across-time, Classification across hospital-services.

show abstract

Application of visual mechanical signal detection and loading platform with super‐resolution based on deep learning

Ding

Zhao

Zhang

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

A visual mechanical signal detection and loading platform with super-resolution based on deep learning is designed to improve the detection accuracy of mechanical signals. The visual mechanical signal detection and loading platform with super-resolution include three-dimensional (3D) biological force quantitative detection platform and the mechanical signal loading platform with 3D magnetic distortion and ultrahigh resolution. In the 3D biological force quantitative detection platform, four 3D force sensors are used to collect mechanical signals, and the improved fuzzy clustering fusion method is used to fuse the mechanical signals collected by 3D force sensors to improve the detection accuracy of mechanical signals. The mechanical signal loading platform of 3D magnetic distortion and ultrahigh resolution technology connects the 3D magnetic distortion instrument and microscope, collects images through high-speed scanning components and distorted magnetic field, reconstructs the collected images by deep learning method, obtains ultrahigh-resolution mechanical signal visual images, and triggers mechanical signal loading and release by synchronous interactive system. The consequences of the experiment demonstrate that the designed platform can display the super-

show abstract

Natural language processing of electronic health records is superior to billing codes to identify symptom burden in hemodialysis patients

Cited by 32 publications

References 28 publications

Considerations for advancing nephrology research and practice through natural language processing

Considerations for advancing nephrology research and practice through natural language processing

Extreme Multi-Label ICD Classification: Sensitivity to Hospital Service and Time

Application of visual mechanical signal detection and loading platform with super‐resolution based on deep learning

Contact Info

Product

Resources

About