Relative apparent diffusion coefficient: a promising tool to differentiate metastatic from benign lymph nodes in animal models.

BACKGROUND

Diffusion-weighted imaging has been widely used to differentiate the character of lymphadenopathy. But there are significant differences between prior studies. The aim of the study was to compare the benefit of apparent diffusion coefficient (ADC) and rADC in the differentiation of metastatic and benign lymph nodes in a rabbit model.

METHODS

Two observers independently measured ADCs in quadriceps of every rabbit on diffusion-weighted images with different sizes of regions of interest (ROI). The appropriate ROI for rADC was determined using the interobserver coefficient. rADC was calculated by ADC(lesion)/ADC(reference site). Receiver operating characteristics (ROC) analysis was performed to compare the diagnostic value of the ADC values and rADC values in differentiating metastatic from benign lymph nodes.

RESULTS

When the ROIs included five pixels, the ICC was 0.816, indicating a good interobserver agreement. The differences of ADC and rADC values between metastatic and benign lymph nodes were both statistically significant. The area under the ROC curve was greater for the rADC than for the ADC. With the rADC criteria of 0.640, the sensitivity and specificity for differentiating metastatic from benign lymph nodes were 93% and 90%, respectively.

CONCLUSION

As a reference site, muscles reveal a good repeatability using a small ROI and the rADC may improve diagnostic accuracy for detecting metastatic nodes in animal models.