• Development and Validation of...

Treffer: Development and Validation of a Protein Electrophoresis Classification Algorithm: Tabular Data-Based Alternative.

Title:
Development and Validation of a Protein Electrophoresis Classification Algorithm: Tabular Data-Based Alternative.
Authors:
Mazuir A; Laboratoire B2A, Brumath, France.; Institut de Recherche Mathématique Avancée, Strasbourg, France., Ricotier G; Laboratoire B2A, Brumath, France., Filhine-Tresarrieu P; Laboratoire B2A, Brumath, France.
Source:
JMIR formative research [JMIR Form Res] 2026 Jan 28; Vol. 10, pp. e83124. Date of Electronic Publication: 2026 Jan 28.
Publication Type:
Journal Article; Validation Study
Language:
English
Journal Info:
Publisher: JMIR Publications Country of Publication: Canada NLM ID: 101726394 Publication Model: Electronic Cited Medium: Internet ISSN: 2561-326X (Electronic) Linking ISSN: 2561326X NLM ISO Abbreviation: JMIR Form Res Subsets: MEDLINE
Imprint Name(s):
Original Publication: Toronto, ON, Canada : JMIR Publications, [2017]-
MeSH Terms:
Blood Protein Electrophoresis*/methods , Blood Proteins*/analysis , Algorithms*, Humans ; Machine Learning ; Deep Learning ; Classification Algorithms
References:
Blood Rev. 2007 Sep;21(5):255-65. (PMID: 17367905)
Clin Chem. 2021 Oct 1;67(10):1406-1414. (PMID: 34491313)
PLoS One. 2022 Aug 24;17(8):e0273284. (PMID: 36001575)
Clin Chem Lab Med. 2024 Jun 17;62(12):2498-2506. (PMID: 38879789)
Contributed Indexing:
Keywords: CatBoost; clinical informatics; computational efficiency; convolutional neural network; diagnostic interpretation; machine learning; serum protein electrophoresis; tabular data analysis
Substance Nomenclature:
0 (Blood Proteins)
Entry Date(s):
Date Created: 20260128 Date Completed: 20260128 Latest Revision: 20260215
Update Code:
20260215
PubMed Central ID:
PMC12895147
DOI:
10.2196/83124
PMID:
41605495
Database:
MEDLINE

Weitere Informationen

Serum protein electrophoresis (SPE) is routinely interpreted through visual assessment of electropherogram images by medical laboratory scientists. We introduce an efficient tabular data-based machine learning approach that directly leverages numerical SPE profiles, offering a robust and interpretable alternative to image-based deep learning methods.
(©Auriane Mazuir, Gatien Ricotier, Pierre Filhine-Tresarrieu. Originally published in JMIR Formative Research (https://formative.jmir.org), 28.01.2026.)