Korean J Radiol.  2016 Oct;17(5):650-656. 10.3348/kjr.2016.17.5.650.

Diffusion Weighted Imaging for Differentiating Benign from Malignant Orbital Tumors: Diagnostic Performance of the Apparent Diffusion Coefficient Based on Region of Interest Selection Method

• Xu XQ
• Hu H
• Su GY
• Liu H
• Shi HB
• Wu FY

• KMID: 2458056
• DOI: http://doi.org/10.3348/kjr.2016.17.5.650

Abstract

OBJECTIVE
To evaluate the differences in the apparent diffusion coefficient (ADC) measurements based on three different region of interest (ROI) selection methods, and compare their diagnostic performance in differentiating benign from malignant orbital tumors.
MATERIALS AND METHODS
Diffusion-weighted imaging data of sixty-four patients with orbital tumors (33 benign and 31 malignant) were retrospectively analyzed. Two readers independently measured the ADC values using three different ROIs selection methods including whole-tumor (WT), single-slice (SS), and reader-defined small sample (RDSS). The differences of ADC values (ADC-ROI(WT), ADC-ROI(SS), and ADC-ROI(RDSS)) between benign and malignant group were compared using unpaired t test. Receiver operating characteristic curve was used to determine and compare their diagnostic ability. The ADC measurement time was compared using ANOVA analysis and the measurement reproducibility was assessed using Bland-Altman method and intra-class correlation coefficient (ICC).
RESULTS
Malignant group showed significantly lower ADC-ROI(WT), ADC-ROI(SS), and ADC-ROI(RDSS) than benign group (all p < 0.05). The areas under the curve showed no significant difference when using ADC-ROI(WT), ADC-ROI(SS), and ADC-ROI(RDSS) as differentiating index, respectively (all p > 0.05). The ROI(SS) and ROI(RDSS) required comparable measurement time (p > 0.05), while significantly shorter than ROIWT (p < 0.05). The ROI(SS) showed the best reproducibility (mean difference ± limits of agreement between two readers were 0.022 [-0.080-0.123] × 10(-3) mm2/s; ICC, 0.997) among three ROI methods.
CONCLUSION
Apparent diffusion coefficient values based on the three different ROI selection methods can help to differentiate benign from malignant orbital tumors. The results of measurement time, reproducibility and diagnostic ability suggest that the ROI(SS) method are potentially useful for clinical practice.

Keyword

Diffusion weighted imaging; Apparent diffusion coefficient; Orbit; Tumor; DWI; ADC

MeSH Terms

Adolescent
Adult
Aged
Aged, 80 and over
Diagnosis, Differential
Diffusion Magnetic Resonance Imaging/*methods
Female
Humans
Image Interpretation, Computer-Assisted/methods
Male
Middle Aged
Observer Variation
Orbital Neoplasms/*diagnostic imaging
ROC Curve
Reproducibility of Results
Retrospective Studies
Sensitivity and Specificity
Young Adult

Figure

• 

  Fig. 1 Schematic illustration of three ROI selection methods on DW (b = 800 s/mm2) images obtained from 70-year-old patient with orbital lymphoma. Axial T2-weighted image shows lesion molding in right orbit (A). For reader-defined small sample ROI, three freehand circular ROIs with mean area of approximately 0.5 cm2 were drawn on area in which tumor showed relative higher signal intensity on DW image (b = 800) (B). For selected slice ROI, slice in which tumor showed largest diameter was chosen. Free-hand ROIs was drawn to cover as much tumor tissue as possible in this slice (C), then corresponding ADC map could be generated and embed within DW image (D). For whole-tumor ROI, freehand ROIs were drawn along border of tumor to cover entire tumor area on each tumor-containing slice (E, four maps). ADC = apparent diffusion coefficient, DW = diffusion weighted, ROI = regions of interest

• 

  Fig. 2 Box-and-whisker plots of ADC-ROIWT (A), ADC-ROISS (B), and ADC-ROIRDSS (C) values between benign and malignant groups. *Outliers. ADC = apparent diffusion coefficient, ROI = region of interest, ROIRDSS = ROI based on reader-defined small sample in selected slices, ROISS = ROI based on single slice, ROIWT = ROI based on whole tumor

• 

  Fig. 3 ROC analyses using ADC-ROIWT (A), ADC-ROISS (B), and ADC-ROIRDSS (C) values to differentiate benign from malignant orbital tumors. Multiple ROC curves comparison indicated no significant differences on area under curves when using ADC-ROIWT, ADC-ROISS, and ADC-ROIRDSS as differentiating index, respectively (ADC-ROIWT vs. ADC-ROISS, p = 0.073; ADC-ROISS vs. ADC-ROIRDSS, p = 0.610; ADC-ROIWT vs. ADC-ROIRDSS, p = 0.064). ADC = apparent diffusion coefficient, AUC = area under curve, ROC = receiver operating characteristic, ROI = region of interest, ROIRDSS = ROI based on reader-defined small sample in selected slices, ROISS = ROI based on single slice, ROIWT = ROI based on whole tumor

• 

  Fig. 4 Bland-Altman plots of inter-reader agreement for ADC-ROIWT (A), ADC-ROISS (B), and ADC-ROIRDSS (C). Difference of mean ADC value between two readers (y-axis) was plotted against mean ADC value of two readers (x-axis), with mean absolute difference (bias) (solid line) and 95% confidence interval of mean difference (limits of agreement) (dashed lines). ADC = apparent diffusion coefficient, ROI = region of interest, ROIRDSS = ROI based on reader-defined small sample in selected slices, ROISS = ROI based on single slice, ROIWT = ROI based on whole tumor

Reference

1. Tailor TD, Gupta D, Dalley RW, Keene CD, Anzai Y. Orbital neoplasms in adults: clinical, radiologic, and pathologic review. Radiographics. 2013; 33:1739–1758.

2. Gufler H, Preiß M, Koesling S. Visibility of sutures of the orbit and periorbital region using multidetector computed tomography. Korean J Radiol. 2014; 15:802–809.

3. Xian J, Zhang Z, Wang Z, Li J, Yang B, Man F, et al. Value of MR imaging in the differentiation of benign and malignant orbital tumors in adults. Eur Radiol. 2010; 20:1692–1702.

4. Lim HK, Lee JH, Baek HJ, Kim N, Lee H, Park JW, et al. Is diffusion-weighted MRI useful for differentiation of small non-necrotic cervical lymph nodes in patients with head and neck malignancies? Korean J Radiol. 2014; 15:810–816.

5. Xu XQ, Hu H, Su GY, Zhang L, Liu H, Hong XN, et al. Orbital indeterminate lesions in adults: combined magnetic resonance morphometry and histogram analysis of apparent diffusion coefficient maps for predicting malignancy. Acad Radiol. 2016; 23:200–208.

6. Sepahdari AR, Kapur R, Aakalu VK, Villablanca JP, Mafee MF. Diffusion-weighted imaging of malignant ocular masses: initial results and directions for further study. AJNR Am J Neuroradiol. 2012; 33:314–319.

7. Xu XQ, Hu H, Su GY, Liu H, Hong XN, Shi HB, et al. Utility of histogram analysis of ADC maps for differentiating orbital tumors. Diagn Interv Radiol. 2016; 22:161–167.

8. Razek AA, Elkhamary S, Mousa A. Differentiation between benign and malignant orbital tumors at 3-T diffusion MR-imaging. Neuroradiology. 2011; 53:517–522.

9. Sepahdari AR, Aakalu VK, Setabutr P, Shiehmorteza M, Naheedy JH, Mafee MF. Indeterminate orbital masses: restricted diffusion at MR imaging with echo-planar diffusion-weighted imaging predicts malignancy. Radiology. 2010; 256:554–564.

10. Haradome K, Haradome H, Usui Y, Ueda S, Kwee TC, Saito K, et al. Orbital lymphoproliferative disorders (OLPDs): value of MR imaging for differentiating orbital lymphoma from benign OPLDs. AJNR Am J Neuroradiol. 2014; 35:1976–1982.

11. Ma C, Liu L, Li J, Wang L, Chen LG, Zhang Y, et al. Apparent diffusion coefficient (ADC) measurements in pancreatic adenocarcinoma: a preliminary study of the effect of region of interest on ADC values and interobserver variability. J Magn Reson Imaging. 2016; 43:407–413.

12. Wu CJ, Wang Q, Li H, Wang XN, Liu XS, Shi HB, et al. DWI-associated entire-tumor histogram analysis for the differentiation of low-grade prostate cancer from intermediate-high-grade prostate cancer. Abdom Imaging. 2015; 40:3214–3221.

13. Colagrande S, Pasquinelli F, Mazzoni LN, Belli G, Virgili G. MR-diffusion weighted imaging of healthy liver parenchyma: repeatability and reproducibility of apparent diffusion coefficient measurement. J Magn Reson Imaging. 2010; 31:912–920.

14. Mukuda N, Fujii S, Inoue C, Fukunaga T, Tanabe Y, Itamochi H, et al. Apparent diffusion coefficient (ADC) measurement in ovarian tumor: effect of region-of-interest methods on ADC values and diagnostic ability. J Magn Reson Imaging. 2016; 43:720–725.

15. Liu L, Ma C, Li J, Wang L, Chen LG, Zhang Y, et al. Comparison of the diagnostic performances of three techniques of ROI placement for ADC measurements in pancreatic adenocarcinoma. Acad Radiol. 2015; 22:1385–1392.

16. Giannotti E, Waugh S, Priba L, Davis Z, Crowe E, Vinnicombe S. Assessment and quantification of sources of variability in breast apparent diffusion coefficient (ADC) measurements at diffusion weighted imaging. Eur J Radiol. 2015; 84:1729–1736.

17. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988; 44:837–845.

18. Cohen J. Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull. 1968; 70:213–220.

19. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1:307–310.

20. Ahlawat S, Khandheria P, Del Grande F, Morelli J, Subhawong TK, Demehri S, et al. Interobserver variability of selective region-of-interest measurement protocols for quantitative diffusion weighted imaging in soft tissue masses: comparison with whole tumor volume measurements. J Magn Reson Imaging. 2016; 43:446–454.

21. Wan H, Sha Y, Zhang F, Hong R, Tian G, Fan H. Diffusion-weighted imaging using readout-segmented echo-planar imaging, parallel imaging, and two-dimensional navigator-based reacquisition in detecting acute optic neuritis. J Magn Reson Imaging. 2016; 43:655–660.