Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 2, 2026Discover OncologyOpen Access

Development and validation of a deep learning model for breast cancer prognosis using the SEER database

View Full Paper
Ask AI
Bookmark
Share

Authors

XLXuanzi LiSZShuyuan ZhangSYShuai Yang

Discussion

Loading...

Member takes

Overview

Retrospective cohort study reveals superior survival prediction by deep learning in breast invasive ductal carcinoma, highlighting potential for personalized treatment planning.

Key Points

  • To develop and validate a DeepSurv-based deep learning model for individualized survival prediction and treatment decision support in patients with breast invasive ductal carcinoma.
  • Identified N=110,346 patients diagnosed with breast invasive ductal carcinoma between 2012 and 2015 from the SEER database, randomly allocating them into training and testing cohorts at a 7:3 ratio.
  • Benchmarked the DeepSurv model against Cox proportional hazards (CPH), Lasso, random survival forests (RSF), XGBoost, and standard clinical staging using C-index, time-dependent ROC curves, Brier scores, integrated Brier score (IBS), and decision curve analysis (DCA).
  • The DeepSurv model achieved a C-index of 0.826 for overall survival (OS) and 0.861 for disease-specific survival (DSS), outperforming CPH (0.814/0.855), Lasso (0.811/0.854), RSF (0.818/0.858), XGBoost (0.799/0.842), and clinical staging (0.708/0.805).
  • Model evaluation showed 3-year and 5-year AUC values of 0.868 and 0.855 for OS (IBS: 0.142) and 0.901 and 0.889 for DSS (IBS: 0.118), alongside favorable net clinical benefit on DCA.
  • Patients receiving treatment concordant with model recommendations demonstrated significantly higher OS and DSS compared to those receiving discordant treatment (both P < 0.001).

Cite This Study

Li et al. (2026) studied this question.

synapsesocial.com/papers/6a97e318c562ede874ec7bf6https://doi.org/10.1007/s12672-026-05848-7
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Deep Learning Models for Predicting the Survival of Patients with Hepatocellular Carcinoma Based on a Surveillance, Epidemiology, and End Results (SEER) Database Analysis2024
  2. 2Deep learning models for predicting the survival of patients with medulloblastoma based on a surveillance, epidemiology, and end results analysis2024 · 1 citations
  3. 3Development and validation of esophageal cancer prediction model based on DeepSurv model: A cohort study2025 · 2 citations
  4. 4Development and external validation of multiple machine learning-based models for breast cancer-specific survival prediction in postoperative patients with invasive breast cancer: a study based on the SEER database and an external cohort2026
  5. 5Deep learning for survival prediction in triple-negative breast cancer: development and validation in real-world cohorts2025 · 6 citations