loading page

Model Generalizability Investigation for GFCE-MRI Synthesis in Radiotherapy of NPC patients Using Multi-institutional Data and Patient-based Data Normalization
  • +7
  • Wen Li ,
  • Saikit Lam ,
  • Tian Li ,
  • Jens Kleesiek ,
  • Andy Lai-Yin Cheung ,
  • Ying Sun ,
  • Francis Kar-ho Lee ,
  • Kwok-hung Au ,
  • Victor Ho Fun Lee ,
  • Jing Cai
Wen Li
The Hong Kong Polytechnic University, The Hong Kong Polytechnic University

Corresponding Author:[email protected]

Author Profile
Saikit Lam
Author Profile
Jens Kleesiek
Author Profile
Andy Lai-Yin Cheung
Author Profile
Francis Kar-ho Lee
Author Profile
Kwok-hung Au
Author Profile
Victor Ho Fun Lee
Author Profile


In this study, we aimed at investigating generalizability of GFCE-MRI model using data from seven institutions by manipulating heterogeneity of training MRI data under two popular normalization approaches. A multimodality-guided synergistic neural network (MMgSN-Net) was applied to map from T1-weighted and T2-weighted MRI to contrast-enhanced MRI (CE-MRI) for GFCE-MRI synthesis in patients with nasopharyngeal carcinoma. MRI data from three institutions were used separately to generate three uni-institution models and jointly for a tri-institution model. Min-Max and Z-Score were applied for data normalization of each model. MRI data from the remaining four institutions served as external cohorts for model generalizability assessment. Quality of GFCE-MRI was quantitatively evaluated against ground-truth CE-MRI using mean absolute error (MAE) and peak signal-to-noise ratio (PSNR). Results showed that performance of all uni-institution models remarkably dropped on the external cohorts. By contrast, model trained using multi-institutional data with Z-Score normalization yielded significantly improved model generalizability
2023Published in IEEE Journal of Biomedical and Health Informatics on pages 1-11. 10.1109/JBHI.2023.3308529