EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Leveraging transfer learning for predicting acute graft- versus -host disease in stem cell transplantation

Lazaro Gibert-Garcia, Rodrigo Guerra, Jocelyn Dunstan, Julia Palma, Axel J. Soto, Ana G. Maguitman, Carlos Chesñevar

PeerJ Computer Science · 19.11.2026 · 12 · e4163

DOI: 10.7717/peerj-cs.4163 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PeerJ Computer Science, online erschienen am 19.11.2026. Verfasst von Lazaro Gibert-Garcia und 6 weiteren Autor:innen. Die Arbeit nennt 76 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: CONICET, MinCyT, Universidad Nacional del Sur.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Acute graft- versus -host disease (aGvHD) is a major complication after allogeneic hematopoietic stem cell transplantation (HSCT), occurring when transplanted donor immune cells attack the recipient’s tissues. Despite advances in HSCT, no reliable clinical method currently exists to predict which patients will develop aGvHD. Early identification of high-risk individuals could improve donor selection and guide preventive strategies. Although machine learning techniques have demonstrated promising predictive performance, their practical deployment is limited by data privacy regulations, heterogeneity in healthcare records, and the scarcity of labeled samples across institutions. We propose a machine learning framework for aGvHD prediction that avoids raw-data sharing and addresses these limitations in decentralized data environments. The framework leverages three publicly available datasets and applies a transfer learning strategy. A source model is trained locally using only shared attributes, and its learned representations are transferred to enhance a target model trained on a different dataset. This approach enables knowledge integration across institutions without sharing sensitive patient-level data. Experimental results demonstrate that the proposed method outperforms baseline models, achieving an area under the receiver operating characteristic curve (AUROC) greater than 0.7 on the three datasets. Statistical significance was assessed using Dunnett’s test and bootstrap confidence interval test, revealing significant improvements in AUROC compared to the majority of baselines ( p < 0.01). A post-hoc explainability analysis further identifies key factors associated with aGvHD development. The findings of this study demonstrate that this strategy yields noteworthy outcomes, especially in terms of AUROC, despite the differences in the diagnoses, donor types, and graft manipulation between the datasets.

Förderung

  • CONICET
  • MinCyT
  • Universidad Nacional del Sur
  • AC3E
  • Fondecyt
  • Cátedra de Transparencia Digital Sostenible Entel, and STIC-AmSud 22-STIC-10

Zitieren

APA

Gibert-Garcia, L., Guerra, R., Dunstan, J., Palma, J., Soto, A. J., Maguitman, A. G., & Chesñevar, C. (2026). Leveraging transfer learning for predicting acute graft- versus -host disease in stem cell transplantation. PeerJ Computer Science, 12, e4163. https://doi.org/10.7717/peerj-cs.4163

Vancouver

Gibert-Garcia L, Guerra R, Dunstan J, Palma J, Soto AJ, Maguitman AG, et al. Leveraging transfer learning for predicting acute graft- versus -host disease in stem cell transplantation. PeerJ Computer Science. 2026;12:e4163. doi:10.7717/peerj-cs.4163

BibTeX
@article{GibertGarcia2026,
  author = {Gibert-Garcia, Lazaro and Guerra, Rodrigo and Dunstan, Jocelyn and Palma, Julia and Soto, Axel J. and Maguitman, Ana G. and Chesñevar, Carlos},
  title = {Leveraging transfer learning for predicting acute graft- versus -host disease in stem cell transplantation},
  journal = {PeerJ Computer Science},
  year = {2026},
  volume = {12},
  pages = {e4163},
  doi = {10.7717/peerj-cs.4163}
}
Kostenlos starten →