Gastric cancer survival prediction using artificial intelligence models based on electronic health records: a systematic review and meta-analysis - Takeaways - MDSpire

Gastric cancer survival prediction using artificial intelligence models based on electronic health records: a systematic review and meta-analysis

  • By

  • Maryana Mandrina

  • Tigran Gevorkyan

  • Sergey Zvezda

  • Valeria Pavlova

  • Rukiyat Abdulaeva

  • Mariam Manukyan

  • Yana Belenkaya

  • Sergey Gordeyev

  • Ivan Stilidi

  • June 23, 2026

  • 0 min

Share

  • 1

    This systematic review evaluates AI models using electronic health records to predict 5-year overall survival in gastric cancer patients post-surgery.

  • 2

    Ten retrospective studies involving 15,643 patients were analyzed, showing a statistically significant AUC improvement of 0.04 for machine learning models.

  • 3

    Boosting techniques outperformed bagging strategies in predictive efficacy, with an AUC improvement of 0.02, indicating algorithm performance variability.

  • 4

    Prognostic factors consistently identified included age, T stage, tumor size, serum albumin levels, and the metastatic-to-examined lymph node ratio.

  • 5

    The study highlights the need for further research on data characteristics that enhance the predictive accuracy of AI-driven prognostic frameworks.

Original Source(s)

Gastric cancer survival prediction using artificial intelligence models based on electronic health records: a systematic review and meta-analysis

  • by Maryana Mandrina, Tigran Gevorkyan, Sergey Zvezda, Valeria Pavlova, Rukiyat Abdulaeva, Mariam Manukyan, Yana Belenkaya, Sergey Gordeyev, Ivan Stilidi

  • June 23, 2026

Related Content

A case report: Retroperitoneal dedifferentiated liposarcoma with heterologous myogenic differentiation and testicular metastasis

Virtual Staining and the Next Era of Pathology

How AI enables scalable biomarker analysis that preserves tissue samples, addressing some of the most persistent bottlenecks in clinical care and research

Editorial: Precision oncology: integrating molecular mechanisms, organoid models, and omics technologies for personalized cancer care