Improving Readability of Swedish Electronic Health Records through Lexical Simplification: First Results

Question

1. What have the authors contributed in "Improving readability of swedish electronic health records through lexical simplification: first results" ?

2. What are the future works in "Improving readability of swedish electronic health records through lexical simplification: first results" ?

3. What is the approach to dealing with unknown words?

4. What is the urgent area for future improvement of the method?

Accepted Answer

This paper describes part of an ongoing effort to improve the readability of Swedish electronic health records ( EHRs ).. The authors report results on lexical simplification of Swedish EHRs, by which they mean detecting the unknown, out-ofdictionary words and trying to resolve them either as compounded known words, abbreviations or misspellings.

Accepted Answer

This distribution result is an important finding for future studies in lexical simplification and readability studies of EHRs, as it might be useful for informing automatic processing approaches.. Second, as a future guideline for the overall pipeline for detecting and resolving unknown, outof-dictionary words, the authors suggest handling abbreviations in a first step, and then taking care of misspellings and potential compounds.

Accepted Answer

Their approach to dealing with unknown words combines a rule-based abbreviation detection and Swedish statistical language model-based compound analysis and misspelling resolution.

Accepted Answer

The most urgent area for future improvement of the method is to handle compound words containing both abbreviations and misspellings.

Accepted Answer

Their approach to clinical Swedish misspellings is based on the best practices of spell checkers for Indo-European languages, namely the phonetic similarity key method combined with a method to measure proximity between the strings.

Accepted Answer

The low recall is due to the design of the SCATM which does not handle words longer than 6 characters, thus resulting in compounded abbreviations like kärlkir or övervak to go undetected.

Accepted Answer

Because many medical terms are multiword expressions that are repeated frequently in a patient’s EHR, the use of acronyms is very common.

Accepted Answer

The most common ap-proach to compound splitting is to base it on a lexicon providing restrictions on how different word forms can be used for generating compounds.

Accepted Answer

The 433 sentences contained a total of 550 abbreviations (78 of these were constituents of compound words), and 41 misspellings of which 13 were misspelled words containing abbreviations.

Accepted Answer

An example of spelling correction of clinical text for other languages is Tolentino et al. (2007), who use several algorithms for word similarity detection, including phonological homonym lookup and ngrams for contextual disambiguation.

Accepted Answer

That is, first, all known abbreviations are detected and marked; second the unknown words are checked whether they are compounds; finally, for the remaining unknown words, context dependent word corrections are made.

Accepted Answer

A major reason for the obstacles to readability of EHRs for laymen stems from the fact that they74are written under time pressure by professionals, for professionals (Kvist et al., 2011).

Accepted Answer

5Approved by the Regional Ethical Review Board in Stockholm (Etikprövningsnämnden i Stockholm), permission number 2012/2028-31/5Normalization of lexis in clinical text relies heavily on the lookup in available lexicons, corpora and domain terminologies.

Accepted Answer

An example of term detection in English clinical texts is Wang and Patrick (2009) work based on rule-based and machine learning methods, reporting 84% precision.

Accepted Answer

This distribution result is an important finding for future studies in lexical simplification and readability studies of EHRs, as it might be useful for informing automatic processing approaches.

Accepted Answer

With this kind of approach, compound splitting is usually very reliable for genres like newspaper text, with typical accuracies for Swedish around 97%, but performs poorer in domain specific genres.

Improving Readability of Swedish Electronic Health Records through Lexical Simplification: First Results

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What have the authors contributed in "Improving readability of swedish electronic health records through lexical simplification: first results" ?

2. What are the future works in "Improving readability of swedish electronic health records through lexical simplification: first results" ?

3. What is the approach to dealing with unknown words?

4. What is the urgent area for future improvement of the method?

5. What is the approach to spelling correction in Swedish?

6. Why did the first experiment achieve low recall?

7. What is the meaning of acronyms in medical terminology?

8. What is the common ap-proach to compound splitting?

9. How many misspellings were found in the sample?

10. What is the common approach to spelling correction of Swedish clinical text?

11. What is the method used to detect and mark abbreviations?

12. What is the main reason for the difficulties to readability of EHRs for laymen?

13. What is the author's opinion on the use of lexis in clinical text?

14. How does Wang and Patrick (2009) use machine learning to detect terms?

15. What is the significance of this distribution result?

16. What is the way to use compound splitting?

Figures

Citations

On the creation of a clinical gold standard corpus in Spanish

HEALTH BANK - A Workbench for Data Science Applications in Healthcare

Learning to extract adverse drug reaction events from electronic health records in Spanish

Artificial Intelligence in Emergency Medicine: Viewpoint of Current Applications and Foreseeable Opportunities and Challenges

Using natural language processing and machine learning to classify health literacy from secure messages: The ECLIPPSE study.

References

Binary codes capable of correcting deletions, insertions, and reversals

Binary codes capable of correcting deletions, insertions and reversals

The Art in Computer Programming

Binary Codes capable of currecting deletions, insertions, and reversals

The Art of Computer Programming, Volume 3, Sorting and Searching

Related Papers (5)

SCAN: A Swedish clinical abbreviation normalizer - Further development and adaptation to radiology

Initial Results in the Development of SCAN A Swedish Clinical Abbreviation Normalizer

Corpus-Driven Terminology Development: Populating Swedish SNOMED CT with Synonyms Extracted from Electronic Health Records

Problems of the English Abbreviations in Medical Translation

Identifying synonymy between SNOMED clinical terms of varying length using distributional analysis of electronic health records.