Crosslingual named entity recognition for clinical de-identification applied to a COVID-19 Italian data set

Abstract

The COrona VIrus Disease 19 (COVID-19) pandemic required the work of all global experts to tackle it. Despite the abundance of new studies, privacy laws prevent their dissemination for medical investigations: through clinical de-identification, the Protected Health Information (PHI) contained therein can be anonymized so that medical records can be shared and published. The automation of clinical de-identification through deep learning techniques has proven to be less effective for languages other than English due to the scarcity of data sets. Hence a new Italian de-identification data set has been created from the COVID-19 clinical records made available by the Italian Society of Radiology (SIRM). Therefore, two multi-lingual deep learning systems have been developed for this low-resource language scenario: the objective is to investigate their ability to transfer knowledge between different languages while maintaining the necessary features to correctly perform the Named Entity Recognition task for de-identification. The systems were trained using four different strategies, using both the English Informatics for Integrating Biology & the Bedside (i2b2) 2014 and the new Italian SIRM COVID-19 data sets, then evaluated on the latter. These approaches have demonstrated the effectiveness of cross-lingual transfer learning to de-identify medical records written in a low resource language such as Italian, using one with high resources such as English.

The application of artificial intelligence and data integration in COVID-19 studies: a scoping review

“…In the lone study that integrated heterogeneous data for modeling, Abdalla et al integrated 43 sociodemographic variables from multiple sources (eg, Census Bureau, US Department of Agriculture, Centers for Disease Control and Prevention) and built elastic net models to examine how sociodemographics impacted county-level social distancing ( Table 4 ). 130 Of the remaining studies, 1 used ANN to perform a drive-through mass vaccination simulation, 138 while the other 4 used NLP methods and tools on various research topics, including cross-lingual clinical deidentification in electronic health records (EHRs), 139 dream reports analysis, 140 drug safety analysis by mining the FDA adverse event system, 141 COVID-19 clinical concept (signs and symptoms) identification, and normalization in EHRs. 142 …”

Data driven covid-19 spread prediction based on mobility and mask mandate information

“…Different models have been proposed in the literature for predicting COVID-19 spread such as Fong et al [ 14 , 15 ], Hernandez et al [ 16 ]. It should be noted that Catelli et al [ 10 , 11 ] performed interesting studies on Italy dataset. We compare our results for Italy and Turkey with the results of [ 16 ], who used an ARIMA-based model.…”

Intelligent system for COVID-19 prognosis: a state-of-the-art survey

“…Also, a DL based drug detection pipeline has been applied to intend and produce new drug-like compounds against COVID-19 [ 65 ] respectively. At present, many attempts [ 124 ][ 125 ] have been made to enhance new analytical advances with both ML as well as DL. Some of the examples include: ML-based transmission of SARS-CoV-2 analyzing designs utilizing a CRISPR-based virus recognition system was confirmed with high sensitivity and speed [ 27 ].…”