Purpose: The study aimed to evaluate diffusion kurtosis imaging (DKI) in the differentiation between benign and malignant sinonasal lesions, and to compare the diagnostic performance of DKI with diffusion weighted imaging (DWI).
Materials And Methods: Eight-one patients with solid sinonasal lesions confirmed by surgery and pathology (46 malignant and 35 benign) underwent conventional MRI, DWI, and DKI. DKI was performed employing a 13 extended b-value ranging from 0 to 2500 s/mm . Apparent diffusion coefficient (ADC) from DWI, kurtosis (K), and diffusion coefficient (D) from DKI were measured and compared between two groups.
Results: ADC and D values were significantly lower in the malignant sinonasal lesions than in the benign sinonasal lesions (1.11 ± 0.41 versus 1.58 ± 0.50 × 10 mm /s and 1.45 ± 0.36 versus 2.03 ± 0.49 × 10 mm /s, respectively, both P < 0001). K value was significantly higher in the malignant lesions than in the benign lesions (0.91 ± 0.23 versus 0.57 ± 0.24, P < 0001). The receiver operating characteristic curve analyses yielded a cutoff ADC value of 1.27 × 10 mm /s for differentiating between benign and malignant lesions, with a sensitivity of 69.6%, a specificity of 77.1% and an accuracy of 74.0%; a cutoff D value of 1.75 × 10 mm /s, with a sensitivity of 82.6%, a specificity of 77.1% and an accuracy of 80.2%; a cutoff K value of 0.63 with a sensitivity of 95.7%, a specificity of 77.1% and an accuracy of 87.7%. The area under the curve of K value was significantly larger than that of ADC value (0.875 versus 0.762; P < 0.05).
Conclusion: K value of DKI demonstrates significantly higher accuracy compared with ADC value for the differentiation between benign and malignant sinonasal lesions. DKI may be a noninvasive method to evaluate the sinonasal lesions.
Level Of Evidence: 1 J. MAGN. RESON. IMAGING 2017;45:1446-1454.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1002/jmri.25500 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!