EUVIMEDEUVIMED entdecken
Open Access

Multicenter Development and Validation of a Multimodal Deep Learning Model to Predict Moderate to Severe AKI

Jay L Koyner iD, Jennie Martin iD, Kyle A Carey, John Caskey, Dana P Edelson iD, Anoop Mayampurath iD, Dmitriy Dligach iD, Majid Afshar iD, Matthew M Churpek iD

Clinical Journal of the American Society of Nephrology · 2025-06 · 20 (6) · 766-778

DOI: 10.2215/cjn.0000000695 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Clinical Journal of the American Society of Nephrology, online erschienen am 15.04.2025. Verfasst von Jay L Koyner und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medicine, University of Chicago, Chicago, Illinois. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIDDK NIH HHS, NLM NIH HHS, NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Key Points We developed and validated a multimodal (structured and unstructured data) model to predict moderate to severe AKI using multicenter data. This multimodal AKI risk score accurately identifies patients who will develop stage 2 AKI over 2 days earlier than serum creatinine alone. The multimodal model performed better than a model based solely on structured data and performed similarly during temporal and site-based validation.

Background

Prior models for the early identification of AKI have used structured data ( e.g ., vital signs and laboratory values). We aimed to develop and validate a deep learning model to predict moderate to severe AKI by combining structured data and information from unstructured notes.

Methods

Adults (18 years or older) admitted to the University of Wisconsin (2009–2020) and the University of Chicago Medicine (2016–2022) were eligible for inclusion. Patients were excluded if they had no documented serum creatinine (SCr), ESKD, an admission SCr ≥3.0 mg/dl, developed ≥stage 2 AKI before reaching the wards or intensive care unit, or required dialysis (KRT) within the first 48 hours. Text from unstructured notes was mapped to standardized concept unique identifiers to create predictor variables, and structured data variables were also included. An intermediate fusion deep learning recurrent neural network architecture was used to predict ≥stage 2 AKI within the next 48 hours. This multimodal model was developed in the first 80% of the data and temporally validated in the next 20%.

Results

There were 339,998 admissions in the derivation cohort and 84,581 in the validation cohort, with 12,748 (3%) developing ≥stage 2 AKI. Patients with ≥stage 2 AKI were older, more likely to be male, had higher baseline SCr, and were more commonly in the intensive care unit ( P < 0.001 for all). The multimodal model outperformed a model based only on structured data for all outcomes, with an area under the receiver operating characteristic curve (95% confidence interval) of 0.88 (0.88 to 0.88) for predicting ≥stage 2 AKI and 0.93 (0.93 to 0.94) for receiving KRT. The area under the precision-recall-curve for ≥stage 2 AKI was 0.20. The results were similar during external validation.

Conclusions

We developed and validated a multimodal deep learning model using structured and unstructured data that predicts the development of severe AKI across the hospital stay for earlier intervention.

Schlagwörter

Stichwörter der Autor:innen

AKIAKI and critical careacute kidney failurebiomarkersclinical nephrologydialysiskidney failuremortality riskrenal replacement therapyrisk factors

Autor:innen und Institutionen

  • Department of Medicine, University of Chicago, Chicago, Illinois.
  • Department of Medicine, University of Wisconsin-Madison. Madison, Wisconsin.
  • Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, Wisconsin.
  • Department of Computer Science, Loyola University Chicago, Chicago, Illinois.

ORCID-Profile: Jay L Koyner, Jennie Martin, Dana P Edelson, Anoop Mayampurath, Dmitriy Dligach, Majid Afshar, Matthew M Churpek

Förderung

  • NIDDK NIH HHS
  • NLM NIH HHS
  • NIH HHS
  • National Institute of Diabetes and Digestive and Kidney Diseases
  • National Institutes of Health

Zitieren

APA

Koyner, J. L., Martin, J., Carey, K. A., Caskey, J., Edelson, D. P., Mayampurath, A., Dligach, D., Afshar, M., & Churpek, M. M. (2025). Multicenter Development and Validation of a Multimodal Deep Learning Model to Predict Moderate to Severe AKI. Clinical Journal of the American Society of Nephrology, 20(6), 766-778. https://doi.org/10.2215/cjn.0000000695

Vancouver

Koyner JL, Martin J, Carey KA, Caskey J, Edelson DP, Mayampurath A, et al. Multicenter Development and Validation of a Multimodal Deep Learning Model to Predict Moderate to Severe AKI. Clinical Journal of the American Society of Nephrology. 2025;20(6):766-778. doi:10.2215/cjn.0000000695

BibTeX
@article{Koyner2025,
  author = {Koyner, Jay L and Martin, Jennie and Carey, Kyle A and Caskey, John and Edelson, Dana P and Mayampurath, Anoop and Dligach, Dmitriy and Afshar, Majid and Churpek, Matthew M},
  title = {Multicenter Development and Validation of a Multimodal Deep Learning Model to Predict Moderate to Severe AKI},
  journal = {Clinical Journal of the American Society of Nephrology},
  year = {2025},
  volume = {20},
  number = {6},
  pages = {766--778},
  doi = {10.2215/cjn.0000000695}
}
Kostenlos starten →