Michael Y. Park MD, Presenter: Nothing to Disclose

Byung Gil Choi MD, Abstract Co-Author: Nothing to Disclose

Byung Hoon Ha, Abstract Co-Author: Nothing to Disclose

Using physically impossible negative ADC values to detect and view motion in diffusion-weighted MR imaging is easy and feasible, and can potentially help MR technicians or radiologists decide whether there was any movement causing significant errors in ADC maps.

ADC values are often used to differentiate or detect various disorders, but the calculation of ADC values rely on using at least two diffusion images with different b-values that are in the same position. Therefore movement during the imaging study can significantly alter the ADC values, but it is often difficult for the MR technician or radiologist to know if the patient moved significantly during the study. We present a simple quantitative value based and qualitative visual based method to detect motion during diffusion-weighted MRI.

ADC values must be positive, and negative values are physically impossible. Therefore negative ADC values can be considered due to motion or artifacts. Since workstations usually discard these negative values, we created a custom program that counts these physically impossible negative ADC values and also creates a visual movement ADC map based on these impossible negative ADC values. Two meat phantoms were each imaged using a head coil with a b-value of 0, and were moved and rotated to various degrees and imaged again using a b-value of 0, 400, and 1000. This was repeated to simulate various degrees of movement and also partially repeated again using a breast coil. Diffusion images before and after movement were used to simulate ADC maps created with movement artifacts. The number of impossible ADC values was calculated for these ADC maps along with creation of visual movement ADC maps. A simple threshold based noise filter to the data was also applied to reduce stochastic effects by noise.

There was a variation in the absolute percentage of impossible ADC to total ADC values by coil type, phantom, and b-values, but a strong linear correlation (p < 0.05) between the percentage of impossible ADC to total ADC values and degree of movement was noted. Degree of movement was easy to evaluate on the visual movement ADC maps.

Park, M, Choi, B, Ha, B, Using Physically Impossible Negative ADC Values to Detect and View Motion in Diffusion-weighted MR Imaging.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11007067.html