EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Confidence-based prediction of antibiotic resistance at the patient level

Juan S Inda-Díaz iD, Anna Johnning iD, Magnus Hessel, Anders Sjöberg, Anna Lokrantz, Lisa Helldal, Mats Jirstrand, Lennart Svensson, Erik Kristiansson iD

mBio · 2026

DOI: 10.1128/mbio.03431-25 ↗

Kurz gefasst

Fachartikel in der Zeitschrift mBio, 2026. Verfasst von Juan S Inda-Díaz und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Mathematical Sciences, Chalmers University of Technology and University of Gothenburg, Gothenburg, Sweden. Hauptthemen laut MEDLINE: Anti-Bacterial Agents, Bacteria, Drug Resistance, Bacterial, Bacterial Infections. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Health and Medical Research Council, Vetenskapsrådet, Chalmers Tekniska Högskola.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Rapid and accurate diagnostics of bacterial infections are necessary for efficient treatment of antibiotic-resistant pathogens. Cultivation-based methods, such as antibiotic susceptibility testing (AST), are limited by bacterial growth rates and seldom yield results before treatment needs to start, increasing patient risk and contributing to antibiotic overprescription. Here, we present a deep-learning method that leverages patient data and available AST results to predict antibiotic susceptibilities that have not yet been measured. After training on three million AST results from 30 European countries, the method achieved an average accuracy of 93% across bacterial species and antibiotics. It predicted susceptibility with an average major error rate below 5% for quinolones, cephalosporins, and carbapenems, and below 8% and 14% for aminoglycosides and penicillins, respectively. Furthermore, the model predicted resistance with an average very major error rate below 10% for cephalosporins, carbapenems, and aminoglycosides, but with higher very major error rates for penicillins and quinolones. We combined the method with conformal prediction and demonstrated accurate estimation of the predictive uncertainty at the patient level. Our results suggest that artificial intelligence-based decision support may offer new means to meet the growing burden of antibiotic resistance. IMPORTANCE Improved diagnostic tools are vital for maintaining efficient treatment of antibiotic-resistant bacteria and for reducing antibiotic overconsumption. In our research, we introduce a new deep learning-based method capable of predicting untested antibiotic resistance phenotypes. The method uses transformers, a powerful artificial intelligence (AI) technique that efficiently leverages both antibiotic susceptibility tests (AST) and patient data simultaneously. The model produces predictions that can be used as time- and cost-efficient alternatives to results from cultivation-based diagnostic assays. Significantly, our study highlights the potential of AI technologies to address the increasing prevalence of antibiotic-resistant bacterial infections.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Anti-Bacterial AgentsBacteriaDrug Resistance, BacterialBacterial InfectionsDeep LearningHumansMicrobial Sensitivity TestsPrediction Algorithms

Stichwörter der Autor:innen

antibiotic resistanceantibiotic susceptibility testingconformal predictiondiagnosticstransformers

Substanzen

Anti-Bacterial Agents

Autor:innen und Institutionen

  • Department of Mathematical Sciences, Chalmers University of Technology and University of Gothenburg, Gothenburg, Sweden.
  • Center for Antibiotic Resistance Research (CARe), University of Gothenburg, Gothenburg, Sweden.
  • Centre for Microbiome Research, School of Biomedical Sciences, Queensland University of Technology, Translational Research Institute, Woolloongabba, Australia.
  • Department of Systems and Data Analysis, Fraunhofer-Chalmers Center, Gothenburg, Sweden.
  • Department of Clinical Microbiology, Sahlgrenska University Hospital, Gothenburg, Sweden.
  • Department of Electrical Engineering, Chalmers University of Technology, Gothenburg, Sweden.

ORCID-Profile: Juan S Inda-Díaz, Anna Johnning, Erik Kristiansson

Förderung

  • National Health and Medical Research Council
  • Vetenskapsrådet
  • Chalmers Tekniska Högskola

Zitieren

APA

Inda-Díaz, J. S., Johnning, A., Hessel, M., Sjöberg, A., Lokrantz, A., Helldal, L., Jirstrand, M., Svensson, L., & Kristiansson, E. (2026). Confidence-based prediction of antibiotic resistance at the patient level. mBio. https://doi.org/10.1128/mbio.03431-25

Vancouver

Inda-Díaz JS, Johnning A, Hessel M, Sjöberg A, Lokrantz A, Helldal L, et al. Confidence-based prediction of antibiotic resistance at the patient level. mBio. 2026. doi:10.1128/mbio.03431-25

BibTeX
@article{IndaDaz2026,
  author = {Inda-Díaz, Juan S and Johnning, Anna and Hessel, Magnus and Sjöberg, Anders and Lokrantz, Anna and Helldal, Lisa and Jirstrand, Mats and Svensson, Lennart and Kristiansson, Erik},
  title = {Confidence-based prediction of antibiotic resistance at the patient level},
  journal = {mBio},
  year = {2026},
  doi = {10.1128/mbio.03431-25}
}
Kostenlos starten →