Intravoxel water diffusion heterogeneity MR imaging of nasopharyngeal carcinoma using stretched exponential diffusion model
- 511 Downloads
To determine the utility of stretched exponential diffusion model in characterisation of the water diffusion heterogeneity in different tumour stages of nasopharyngeal carcinoma (NPC).
Materials and methods
Fifty patients with newly diagnosed NPC were prospectively recruited. Diffusion-weighted MR imaging was performed using five b values (0–2,500 s/mm2). Respective stretched exponential parameters (DDC, distributed diffusion coefficient; and alpha (α), water heterogeneity) were calculated. Patients were stratified into low and high tumour stage groups based on the American Joint Committee on Cancer (AJCC) staging for determination of the predictive powers of DDC and α using t test and ROC curve analyses.
The mean ± standard deviation values were DDC = 0.692 ± 0.199 (×10−3 mm2/s) for low stage group vs 0.794 ± 0.253 (×10−3 mm2/s) for high stage group; α = 0.792 ± 0.145 for low stage group vs 0.698 ± 0.155 for high stage group. α was significantly lower in the high stage group while DDC was negatively correlated. DDC and α were both reliable independent predictors (p < 0.001), with α being more powerful. Optimal cut-off values were (sensitivity, specificity, positive likelihood ratio, negative likelihood ratio) DDC = 0.692 × 10−3 mm2/s (94.4 %, 64.3 %, 2.64, 0.09), α = 0.720 (72.2 %, 100 %, −, 0.28).
The heterogeneity index α is robust and can potentially help in staging and grading prediction in NPC.
• Stretched exponential diffusion models can help in tissue characterisation in nasopharyngeal carcinoma
• α and distributed diffusion coefficient (DDC) are negatively correlated
• α is a robust heterogeneity index marker
• α can potentially help in staging and grading prediction
KeywordsNasopharyngeal carcinoma Diffusion weighted imaging Magnetic resonance imaging Stretched exponential Staging
Alpha (intravoxel water diffusion heterogeneity)
American Joint Committee on Cancer
Area under curve
Distributed diffusion coefficient
Intravoxel incoherent motion
Positron emission tomography with computed tomography
Receiver operating characteristic
Spectral presaturation inversion recovery
Short TI inversion recovery
Repetition time/echo time
Turbo spin echo
The scientific guarantor of this publication is Prof. Khong Pek Lan. The authors of this manuscript declare relationships with the following companies: Dr. Q Chan is currently employed by Philips Medial Systems. This study has received funding by University Grants Council (UGC) seed funding from The University of Hong Kong, project no. 201112159010.
No complex statistical methods were necessary for this paper. Institutional review board approval was obtained. Written informed consent was obtained from all subjects (patients) in this study. Approval from the institutional animal care committee was not required because this study did not involve animals. Study subjects or cohorts have not been previously reported.
Methodology: prospective, diagnostic or prognostic study, performed at one institution.