Multitask learning for early treatment response and survival prediction in lung cancer radiotherapy using sequential CBCT imaging - Takeaways - MDSpire

Deep Learning Approaches for Predicting Treatment Response and Survival Outcomes in Lung Cancer Radiotherapy Through Sequential CBCT Imaging

  • By

  • Yumei Li

  • Zhouji Wei

  • Chenlong Luo

  • Yejun Gong

  • Ye Zhang

  • Feng Zhang

  • Chunyan Liu

  • July 21, 2026

Share

  • 1

    A multitask deep learning framework was developed for treatment response classification and progression-free survival prediction in lung cancer patients.

  • 2

    The model was trained on 142 lung cancer patients using planning CT, dose, and sequential CBCT imaging data.

  • 3

    The first on-treatment CBCT achieved an AUC of 0.858 and a C-index of 0.672, outperforming 12 baseline methods.

  • 4

    Including additional CBCT scans degraded classification performance due to increased dimensionality relative to cohort size.

  • 5

    The study highlights the importance of early imaging in predicting treatment response and survival outcomes in lung cancer.

Original Source(s)

Multitask learning for early treatment response and survival prediction in lung cancer radiotherapy using sequential CBCT imaging

  • by Yumei Li, Zhouji Wei, Chenlong Luo, Yejun Gong, Ye Zhang, Feng Zhang, Chunyan Liu

  • July 21, 2026

Related Content

Global research trends and emerging frontiers of intratumoral microbiota in cancer immunotherapy: a bibliometric and visualization analysis

Non-anticoagulant heparins: glycan-mediated immune modulation and therapeutic applications

From patient-derived tumor organoids to personalized cancer therapy: advancing treatment for advanced solid tumors