CodiEsp-abstracs: Abstracts from Lilacs and Ibecs with ICD10 codes

<p>JSON file with abstracts from Lilacs and Ibecs with ICD10 codes (ICD10-CM and ICD10-PCS) associated to them (CIE10 in Spanish).</p> <p>&nbsp;</p> <p><strong>Please, cite us:</strong></p> <p>Miranda-Escalada, A., Gonzalez-Agirre, A., Armengol-Estap&eacute;, J., Krallinger, M.: Overview of automatic clinical coding: annotations, guidelines, and solutions for non-English clinical cases at CodiEsp track of eHealth CLEF 2020.<em>&nbsp;In: CLEF (Working Notes) (2020)</em></p> <pre><code>@inproceedings{miranda2020overview, </code> <code>title={Overview of automatic clinical coding: annotations, guidelines, and solutions for non-english clinical cases at codiesp track of CLEF eHealth 2020}, </code> <code>author={Miranda-Escalada, Antonio and Gonzalez-Agirre, Aitor and Armengol-Estap{\&#39;e}, Jordi and Krallinger, Martin}, </code> <code>booktitle={Working Notes of Conference and Labs of the Evaluation (CLEF) Forum. CEUR Workshop Proceedings}, </code> <code>year={2020} }</code></pre> <p>&nbsp;</p> <p>Lilacs and Ibecs databases&nbsp;have MeSH terms describing&nbsp;some of their documents. Then, using UMLS Metathesaurus, those MeSH terms have been translated into ICD10 codes (ICD10-CM and ICD10-PCS). Every abstract have at least one ICD10 code.&nbsp;</p> <p>In addition, MeSH codes given by the databases (Lilacs and Ibecs) have a &quot;word&quot; describing them. These &quot;words&quot; have been used to add further ICD10 codes. We have done strict string matching to find whether those &quot;words&quot; were a&nbsp;descriptor of any ICD10 code (in the Spanish version, CIE10).</p> <p>The format of the JSON file is the following:</p> <pre>{&#39;articles&#39;: [{&#39;title&#39;: &#39;title&#39;, &#39;pmid&#39;: &#39;pmid&#39;, &#39;abstractText&#39;: &#39;abtract (in Spanish)&#39;, &#39;Mesh&#39;: [{&#39;Code&#39;: &#39;MeSHCode&#39;, &#39;Word&#39;: &#39;reference&#39;, &#39;CIE&#39;: [CIE10_1, CIE10_2, ...]}, ...] }, ...] }</pre> <p>&nbsp;</p> <p>Additionally, the compressed file includes a folder with all the abstracts extracted in individual UTF-8 encoded text files and a tab-separated file with 4 fields:</p> <blockquote> <p>pmid&nbsp;&nbsp; &nbsp;label&nbsp; &nbsp; cie10-code&nbsp;&nbsp; &nbsp;word</p> </blockquote> <p>Summary statistics:</p> <ul> <li>number of abstracts: 355 840</li> <li>number abstracts with at least one ICD10&nbsp;code: 176 294</li> <li>Percentage of MeSH codes mapped to ICD10: 10.6% (there were&nbsp;2 526 772 MeSH codes and&nbsp;266 949 mapped to ICD10)</li> <li>average number of MeSH codes per article: 7.1</li> <li>average number of ICD10 codes per article: 2.5</li> <li>number of ICD10 codes that have an associated MeSH code in UMLS: 3293</li> <li>number of ICD10 codes that have an associated MeSH code in UMLS and appear in this dataset: 3082</li> </ul>