American Association for Cancer Research
crc-22-0152_fig4.png (170.56 kB)

FIGURE 4 from Multi-institutional Prognostic Modeling in Head and Neck Cancer: Evaluating Impact and Generalizability of Deep Learning and Radiomics

Download (170.56 kB)
posted on 2023-06-29, 14:20 authored by Michal Kazmierski, Mattea Welch, Sejin Kim, Chris McIntosh, Katrina Rey-McIntyre, Shao Hui Huang, Tirth Patel, Tony Tadic, Michael Milosevic, Fei-Fei Liu, Adam Ryczkowski, Joanna Kazmierska, Zezhong Ye, Deborah Plana, Hugo J.W.L. Aerts, Benjamin H. Kann, Scott V. Bratman, Andrew J. Hope, Benjamin Haibe-Kains

Top performing model. A, Overview of Deep MTLR. The model combines EMR features with tumor volume using a neural network and learns to jointly predict the probability of death at all intervals on the discretized time axis, allowing it to achieve good performance in both the binarized and lifetime risk prediction tasks. A predicted survival curve can be constructed for each individual to determine the survival probability at any timepoint. B, Importance of combined input data for performance on the binary endpoint. Training the deep MTLR on EMR features only led to notably worse performance. Furthermore, using a deep convolutional neural network in place of tumor volume did not improve the 2-year AUROC.


Canadian HIV Trials Network, Canadian Institutes of Health Research (CTN, CIHR)



ML combined with simple prognostic factors outperformed multiple advanced CT radiomics and deep learning methods. ML models provided diverse solutions for prognosis of patients with HNC but their prognostic value is affected by differences in patient populations and require extensive validation.