Related Experiment Video
Updated: Nov 7, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Radiomics Model to Predict Early Progression of Nonmetastatic Nasopharyngeal Carcinoma after Intensity Modulation
Richard Du1, Victor H Lee1, Hui Yuan1
1Departments of Diagnostic Radiology (R.D., H.Y., P.L.K., V.V.) and Clinical Oncology (V.H.L., K.O.L., A.W.L., D.L.K.) and the School of Public Health (H.H.P.), Li Ka Shing Faculty of Medicine, The University of Hong Kong, Room 406, Block K, Queen Mary Hospital, Pok Fu Lam Road, Hong Kong SAR; Department of Radiology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China (Y.C.); and Department of Electrical and Electronic Engineering, Faculty of Engineering, The University of Hong Kong, Hong Kong SAR (E.Y.L.).
Machine learning models using MRI radiomics can predict disease progression in nasopharyngeal carcinoma (NPC) patients. Key features include tumor shape, MRI characteristics, and clinical staging, aiding treatment assessment.
Area of Science:
- Radiology
- Oncology
- Artificial Intelligence
Background:
- Nasopharyngeal carcinoma (NPC) is a significant health concern.
- Accurate prediction of disease progression is crucial for patient management.
- Intensity-modulated radiation therapy (IMRT) is a standard treatment for NPC.
Purpose of the Study:
- To evaluate the prognostic capability of a machine learning model using pretreatment MRI radiomic features for predicting 3-year disease progression in nonmetastatic NPC patients.
- To identify and explain the radiomics features that are important predictors of disease progression.
Main Methods:
- Retrospective review of 277 nonmetastatic NPC patients.
- Extraction of 525 radiomic features from MRI and 5 clinical features.
- Development of a support vector machine (SVM) model for predicting 3-year disease progression.
- Application of Shapley Additive Explanations (SHAP) for model interpretability.
Main Results:
- The machine learning model achieved an area under the receiver operating characteristic curve (AUC) of 0.80 in the discovery cohort and 0.73-0.89 in the validation cohort.
- SHAP analysis identified tumor shape sphericity, first-order mean absolute deviation, T stage, and overall stage as significant predictors of 3-year disease progression.
Conclusions:
- Radiomics holds significant potential in assessing NPC prognosis.
- Machine learning models, when combined with explainability techniques like SHAP, can elucidate complex feature interactions for better clinical understanding.
- These findings support the integration of radiomics into NPC management strategies.
More Related Videos
07:57Positron Emission Tomography-based Dose Painting Radiation Therapy in a Glioblastoma Rat Model using the Small Animal Radiation Research Platform
Published on: March 24, 2022
10:25Author Spotlight: Integrating High-Resolution Intravital Imaging and MRI to Enhance Stereotactic Body Radiation Therapy Planning
Published on: April 12, 2024