Applications of machine learning methods in long COVID phenotyping: Review
Annales Academiae Medicae Silesiensis · 19.06.2026
DOI: 10.18794/aams/222516 ↗Kurz gefasst
Fachartikel in der Zeitschrift Annales Academiae Medicae Silesiensis, online erschienen am 19.06.2026. Verfasst von Martyna Szlenk und 5 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Introduction
Long COVID (LC) is one of the major challenges of modern medicine. The lack of standardized diagnostic criteria, nonspecific symptoms, their temporal variability and chronic nature significantly hinder diagnosis and classification. This creates the need for advanced technological tools to better understand the pathophysiology and identify individual disease phenotypes. Artificial intelligence, particularly machine learning (ML), is a promising approach. This review synthesises current evidence on ML applications in LC phenotyping and outlines core ML concepts relevant to this research context. Material and methods A search of the PubMed database identified 47 articles published through July 9, 2024. After manual screening of titles and abstracts, 17 studies published in English were included in the final analysis.
Results
We present an overview of ML applications in LC research, focusing on the identification of phenotypes, subphenotypes and risk factors associated with LC occurrence and severity. The identified phenotypes depict LC as a multisystem condition, in which symptom clusters often involve multiple organ systems rather than a single organ.
Conclusions
ML represents a useful tool for identifying LC patients and classifying them into distinct subphenotypes characterized by different clinical manifestations. This approach improves the precision of diagnostic pathways and supports individualized therapeutic strategies. ML approaches may also improve clinical trial design by facilitating analysis of large, heterogeneous datasets – a particular advantage given LC's variable clinical course.
Zitieren
APA
Szlenk, M., Żmudka, K., Spychał, A., Kołodziej, R., Piłat, P., & Jaroszewicz, J. (2026). Applications of machine learning methods in long COVID phenotyping: Review. Annales Academiae Medicae Silesiensis. https://doi.org/10.18794/aams/222516
Vancouver
Szlenk M, Żmudka K, Spychał A, Kołodziej R, Piłat P, Jaroszewicz J. Applications of machine learning methods in long COVID phenotyping: Review. Annales Academiae Medicae Silesiensis. 2026. doi:10.18794/aams/222516
BibTeX
@article{Szlenk2026,
author = {Szlenk, Martyna and Żmudka, Karol and Spychał, Aleksandra and Kołodziej, Rafał and Piłat, Patrycja and Jaroszewicz, Jerzy},
title = {Applications of machine learning methods in long COVID phenotyping: Review},
journal = {Annales Academiae Medicae Silesiensis},
year = {2026},
doi = {10.18794/aams/222516}
}
EUVIMED