📝 Publications

📓 Journal Paper

Medical Physics
sym

Towards reliable head and neck cancers locoregional recurrence prediction using delta-radiomics and learning with rejection option
Kai Wang, Michael Dohopolski, Qiongwen Zhang, David Sher, Jing Wang, Medical physics, 2022.

Project |

  • Delta-radiomcis features calculated based on Pre- and Post-Therapy PET/CT.
  • Multi-modality model for early-stage locoregional recurrence prediction.
  • We performed patient-specific prediction uncertainty estimation for learning with rejection option to improve the model reliability (higher performance on high-confidence group).
The Laryngoscope
sym

Joint Vestibular Schwannoma Enlargement Prediction and Segmentation Using a Deep Multi-task Model
Kai Wang, Nicholas George-Jones, Liyuan Chen, Jacob Hunter, Jing Wang, The Laryngoscope, 2022.

Project |

  • We presented a deep multi-task (DMT) model to predict vestibular schwannoma enlargement and tumor segmentation mask simultaneously using the initial diagnostic ceT1 MRI.
  • The proposed DMT model is of higher learning efficiency and prediction accuracy than single-task prediction model.
Medical Physics
sym

A multi‐objective radiomics model for the prediction of locoregional recurrence in head and neck squamous cell cancer
Kai Wang, Zhiguo Zhou, Rongfang Wang, Liyuan Chen, Qiongwen Zhang, David Sher, Jing Wang, Medical physics 47.10 (2020): 5392-5400.

  • We built a multi-classifier, multi-objective, and multi-modality (mCOM) model for HNSCC LRR prediction.
  • Our study shows fusing multiple classifiers and multiple modalities can improve the robustness of the predictive model. Additionally, the multi-objective model could help to build a predictive model that can be flexibly adapted to different clinical preferences.

📑 Conference Proceeding

HECKTOR 2022
sym

Recurrence-free Survival Prediction under the Guidance of Automatic Gross Tumor Volume Segmentation for Head and Neck Cancers
Kai Wang, Yunxiang Li, Michael Dohopolski, Tao Peng, Weiguo Lu, You Zhang, Jing Wang (The first two authors contributed to this work equally, Oral 🎤)

  • Ranked No.3 in the recurrece-free survival (RFS) prediction task of the head and neck tumor segmentation and outcome prediction challenge (HECKTOR 2022).
  • nnUNet for GTVp and GTVn segmentation on a multi-institution dataset.
  • Extract radiomics feature from the AI-based auto-segmentated ROI for radiomics-based HNC RFS prediction.
  • Ensemble multi-modality model for more robust prediction.

📰 Conference Abstract

📰 Patent

✒️ In Preparation