2017
DOI: 10.1007/978-3-319-66429-3_51
|View full text |Cite
|
Sign up to set email alerts
|

Medical Speech Recognition: Reaching Parity with Humans

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
14
0

Year Published

2018
2018
2021
2021

Publication Types

Select...
3
3
2

Relationship

0
8

Authors

Journals

citations
Cited by 27 publications
(14 citation statements)
references
References 45 publications
0
14
0
Order By: Relevance
“…More recently, a neural network based speech recognition system has been built for the medical domain using relatively small medical speech data (270 hours) and has been benchmarked against medical transcriptionists [3]. Speech recognition systems have been evaluated on a clinical question answering task and it has been shown that domain adaptation with a language model improves the accuracy in interpreting spoken clinical questions significantly [4].…”
Section: Introductionmentioning
confidence: 99%
“…More recently, a neural network based speech recognition system has been built for the medical domain using relatively small medical speech data (270 hours) and has been benchmarked against medical transcriptionists [3]. Speech recognition systems have been evaluated on a clinical question answering task and it has been shown that domain adaptation with a language model improves the accuracy in interpreting spoken clinical questions significantly [4].…”
Section: Introductionmentioning
confidence: 99%
“…46 Another study developed a deep neural network model for medical voice recognition trained on over 270 hours of speech data and compared the performance to professional medical transcriptionists. 47 The model had a 15.4% error rate when applied to a "realistic clinical use case" and performed equally as well as humans. 47 The 2 studies discussed here illustrate the ability of ML to recognize real conversations between patients and providers.…”
Section: Clinical Assistancementioning
confidence: 94%
“…47 The model had a 15.4% error rate when applied to a "realistic clinical use case" and performed equally as well as humans. 47 The 2 studies discussed here illustrate the ability of ML to recognize real conversations between patients and providers. With further refinement, this software could be hugely beneficial for clinical and office work, creating less need for providers to manually input notes.…”
Section: Clinical Assistancementioning
confidence: 94%
“…To date, research effort has focused on solving foundational problems in the development of a digital scribe, including ASR of medical conversations, 10,11 automatically populating the review of symptoms discussed in a medical encounter, 12 extracting symptoms from medical conversations, 13,14 and generating medical reports from dictations. 15,16 While these developments are promising, several challenges hinder the implementation of a fully functioning digital scribe and its evaluation in a clinical environment.…”
Section: Introductionmentioning
confidence: 99%