Electronic health records naturally contain most of the medical information in the form of doctor's notes as unstructured or semi-structured texts. Current deep learning text analysis approaches allow researchers to reveal the inner semantics of text information and even identify hidden consequences that can offer extra decision support to doctors. In the presented article, we offer a new automated analysis of Polish summary texts of patient hospitalizations. The presented models were found to be able to predict the final diagnosis with almost 70% accuracy based just on the patient's medical history (only 132 words on average), with possible accuracy increases when adding further sentences from hospitalization results; even one sentence was found to improve the results by 4%, and the best accuracy of 78% was achieved with five extra sentences. In addition to detailed descriptions of the data and methodology, we present an evaluation of the analysis using more than 50,000 Polish cardiology patient texts and dive into a detailed error analysis of the approach. The results indicate that the deep analysis of just the medical history summary can suggest the direction of diagnosis with a high probability that can be further increased just by supplementing the records with further examination results.

1st Department of Cardiology Medical University of Silesia 40 055 Katowice Poland

Department of Cardiology and Structural Heart Diseases School of Medicine in Katowice Medical University of Silesia 40 055 Katowice Poland

Interventional Cardiac Electrophysiology Group International Clinical Research Center St Anne's University Hospital Brno 656 91 Brno Czech Republic

Natural Language Processing Centre Faculty of Informatics Masaryk University 602 00 Brno Czech Republic

000      

00000naa a2200000 a 4500

001      

bmc22017094

003      

CZ-PrNML

005      

20220720100110.0

007      

ta

008      

220718s2022 sz f 000 0|eng||

009      

AR

024    7_

$a 10.3390/jpm12060869 $2 doi

035    __

$a (PubMed)35743653

040    __

$a ABA008 $b cze $d ABA008 $e AACR2

041    0_

$a eng

044    __

$a sz

100    1_

$a Anetta, Kristof $u Natural Language Processing Centre, Faculty of Informatics, Masaryk University, 602 00 Brno, Czech Republic $1 https://orcid.org/0000000152379735

245    10

$a Deep Learning Analysis of Polish Electronic Health Records for Diagnosis Prediction in Patients with Cardiovascular Diseases / $c K. Anetta, A. Horak, W. Wojakowski, K. Wita, T. Jadczyk

520    9_

$a Electronic health records naturally contain most of the medical information in the form of doctor's notes as unstructured or semi-structured texts. Current deep learning text analysis approaches allow researchers to reveal the inner semantics of text information and even identify hidden consequences that can offer extra decision support to doctors. In the presented article, we offer a new automated analysis of Polish summary texts of patient hospitalizations. The presented models were found to be able to predict the final diagnosis with almost 70% accuracy based just on the patient's medical history (only 132 words on average), with possible accuracy increases when adding further sentences from hospitalization results; even one sentence was found to improve the results by 4%, and the best accuracy of 78% was achieved with five extra sentences. In addition to detailed descriptions of the data and methodology, we present an evaluation of the analysis using more than 50,000 Polish cardiology patient texts and dive into a detailed error analysis of the approach. The results indicate that the deep analysis of just the medical history summary can suggest the direction of diagnosis with a high probability that can be further increased just by supplementing the records with further examination results.

655    _2

$a časopisecké články $7 D016428

700    1_

$a Horak, Ales $u Natural Language Processing Centre, Faculty of Informatics, Masaryk University, 602 00 Brno, Czech Republic $1 https://orcid.org/000000016348109X

700    1_

$a Wojakowski, Wojciech $u Department of Cardiology and Structural Heart Diseases, School of Medicine in Katowice, Medical University of Silesia, 40-055 Katowice, Poland

700    1_

$a Wita, Krystian $u First Department of Cardiology, Medical University of Silesia, 40-055 Katowice, Poland

700    1_

$a Jadczyk, Tomasz $u Department of Cardiology and Structural Heart Diseases, School of Medicine in Katowice, Medical University of Silesia, 40-055 Katowice, Poland $u Interventional Cardiac Electrophysiology Group, International Clinical Research Center, St. Anne's University Hospital Brno, 656 91 Brno, Czech Republic $1 https://orcid.org/0000000279559100

773    0_

$w MED00203320 $t Journal of personalized medicine $x 2075-4426 $g Roč. 12, č. 6 (2022)

856    41

$u https://pubmed.ncbi.nlm.nih.gov/35743653 $y Pubmed

910    __

$a ABA008 $b sig $c sign $y - $z 0

990    __

$a 20220718 $b ABA008

991    __

$a 20220720100106 $b ABA008

999    __

$a ind $b bmc $g 1816438 $s 1168336

BAS    __

$a 3

BAS    __

$a PreBMC

BMC    __

$a 2022 $b 12 $c 6 $e 20220525 $i 2075-4426 $m Journal of personalized medicine $n J. pers. med. $x MED00203320

GRA    __

$a LM2018101 $p Ministry of Education of CR within the LINDAT-CLARIAH-CZ project

GRA    __

$a project MUNI/IGA/1326/2021 $p Grant Agency of Masaryk University

GRA    __

$a PCN-1-005/N/0/K and PCN-1-073/N/1/K $p Medical University of Silesia in Poland

GRA    __

$p Mieczysław Koćwin Foundation Scholarship

LZP    __

$a Pubmed-20220718