Higher Sensitivity and Specificity for Diffusion-weighted Imaging of Malignant Lung Lesions without Apparent Diffusion Coefficient Quantification

Purpose: To compare the performance of apparent diffusion coefficient (ADC) with that of signal intensity of the lesion-tospinal cord ratio (LSR) in differentiating lung cancer from benign lesions on high-b value diffusion-weighted (DW) magnetic resonance (MR) images. Materials and Methods: This study received institutional review board approval; written informed consent was provided by all patients. Twenty-eight patients (six women, 22 men; mean age, 64.2 years) with pulmonary nodules seen at chest computed tomography were prospectively reviewed. Two DW images with different motion-probing gradient strengths (b(h) = 1000 sec/mm(2) and b(1) = 0 sec/mm(2)) were analyzed semiquantitatively by measuring the signal intensities of the lesions and the spinal cord. ADC was calculated by using linear regression analysis of the natural logarithm of mean signal intensity versus the gradient factor. For reference, LSR was also measured on the same image with a diffusion gradient of b(h) = 1000 sec/mm(2). Results: The LSR of cancer nodules was significantly higher than that of benign lesions, while there were no significant differences between them in relation to ADC. In the receiver operating characteristic curve analysis, LSR had a higher area under the curve than did ADC (0.911 vs 0.600). By using a cutoff value of 1.135, LSR had a positive predictive value of 86.7% a negative predictive value of 90%, and an accuracy of 85.7% for the detection of lung cancer with LSR. Conclusion: LSR measurement on high-b value DW imaging may be useful for differentiating between benign and malignant lung nodules. (C) RSNA, 2009