JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Multimodality radiomics for tumor prognosis in nasopharyngeal carcinoma

  • 0Department of Radiological Technology and Medical Physics, Faculty of Allied Health Sciences, Chulalongkorn University, Bangkok, Thailand.
Plos One +

|

February 12, 2024

|

February 12, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

Radiomic features from CT and MR images significantly improve predictions of overall survival, progression-free survival, and distant metastasis-free survival in nasopharyngeal carcinoma (NPC) patients. This approach enhances prognostic accuracy beyond traditional clinical data.

Area Of Science

  • Oncology
  • Radiology
  • Medical Imaging Analysis

Background

  • Nasopharyngeal carcinoma (NPC) prognosis is difficult due to late detection and undetectable Epstein-Barr virus (EBV) DNA.
  • Radiomic features, quantifying tumor imaging characteristics, offer potential for improved NPC prognosis assessment.

Purpose Of The Study

  • To evaluate the predictive capability of radiomic features for overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in NPC.
  • To assess the added value of radiomics in conjunction with clinical data for NPC survival prediction.

Main Methods

  • Retrospective analysis of 183 NPC patients treated with chemoradiotherapy (2010-2019) with a minimum three-year follow-up.
  • Extraction of radiomic features from pretreatment CT and MR images using PyRadiomics v.2.0.
  • Combination of selected radiomic features with clinical data (age, gender, stage, EBV DNA) for prognostic modeling using Cox regression with RFE and cross-validation.

Main Results

  • Integration of radiomics with clinical data significantly improved predictive power (C-index 0.788-0.848) compared to clinical data alone (C-index 0.745-0.766, p<0.05).
  • Multimodality radiomics combined with clinical data demonstrated the highest performance.
  • Radiomics integration enhanced survival predictions even without EBV DNA data (C-index 0.770-0.831 vs. 0.727-0.734, p<0.05).

Conclusions

  • Multimodality radiomic features from CT and MR images provide superior predictive performance for NPC survival outcomes (OS, PFS, DMFS).
  • Radiomics integration offers a valuable tool for enhancing prognostic assessment in NPC, complementing traditional clinical data.