To develop a motion insensitive 3D high resolution DWI technique to improve quantitative diffusion measurements that can assist in the identification of plaque components in the cervical carotid artery.

DWI has the potential to provide complementary information that will allow better discrimination of plaque components. The ADC values in different components within atherosclerotic plaques suggest that DWI might provide a tool for discriminating intraplaque hemorrhage and lipid necrotic core from other components.^1,2 However, the current 3D DWI sequences based on multi-shot techniques are commonly limited by inconstant phase variance between each shot induced by motion.³ To develop a motion insensitive 3D high resolution DWI technique to improve the accuracy of ADC measurements in important plaque components, a 3D diffusion weighted driven equilibrium Stack of Stars (3D DW-DE SOS) sequence was implemented. This pulse sequence is shown in fig 1. The DW-DE consists of an excitation pulse followed by two refocusing pulses and a tip-up pulse. Two refocusing pulses were designed with three composite hard pulses. A total four sets of diffusion sensitized bipolar gradients were applied to reduce the eddy current effect and to compensate motion.⁴ After DW-DE preparation, data was immediately acquired using a 3D segmented FLASH SOS trajectory. Bloch simulation of the sequence found the transverse magnetization after nth excitation pulse was $$M_{t}(n)=M_{1}(n)e^{-bD}+M_{2}(n)$$ where $$M_{1}(n)=M_{0}e^{-\frac{TEp}{T2}}e^{-\frac{d}{T1}}(e^{-\frac{TR}{T1}}\cos\alpha)^{n-1}\sin\alpha$$, M_{2}(n)=[M_{0}(1-e^{-\frac{d}{T1}})(e^{-\frac{TR}{T1}}\cos\alpha)^{n-1}+M_{0}(1-e^{-\frac{TR}{T1}})(\frac{(1-(e^{-\frac{TR}{T1}}\cos\alpha)^{n-1})}{1-e^{-\frac{TR}{T1}}\cos\alpha})]\sin\alpha$$. The transverse magnetization can be separated into a diffusion dependent term $$M_{1}(n)e^{-bD}$$ and diffusion independent term $$M_{2}(n)$$ originating from the T1 relaxation effect. To minimize the effect of T1, following the DW-DE prep and a fat saturation pulse, all kz partitions for the same radial spoke are acquired with centric ordering in the same turbo-FLASH echotrain.⁵ All studies were performed on a Siemens Trio 3T MRI scanner (Siemens Medical Solutions, Erlangen, Germany). The carotid arteries of six patients who showed hemorrhage signal from the previous studies were acquired using 3D DW-DE SOS. The imaging parameters in DWI were: FOV=152x152 mm², 2 mm slice thickness, TE/TR = 2.05/8.0ms, 32 slices/slab, b =20, 450 s/mm². The resultant in-plane spatial resolution was 0.6x0.6 mm². The total imaging time was 3 min 20 sec. 3D MPRAGE and 3D T1 SPACE with DANTE flow suppression sequences were also performed with voxel dimension of 0.73 mm. The ADC map was calculated from b=20, 450s/mm² images and displayed using IDL( Exelis Visual Information Solutions). ADC values were obtained for the normal wall segments on 3 adjacent slices of the common carotid artery on all six subjects. ADC values in the plaque of the 6 subjects were measured at 3 different slice locations within the hemorrhages.

The mean ADC value of vessel walls from all subjects was 1.32±0.29x10^-3mm²/s. Figure 2 displays the 3D T1 DANTE SPACE, b=20 and 450s/mm² images of four contiguous slices from subject 2. As shown by arrows in Fig 2, the DWI including ADC map demonstrates clear contrast between wall (yellow) and plaque (red) area. Fig 3 displays the 3D MPRAGE, 3D T1 SPACE and b=20s/mm² images and ADC maps of one subject with bilateral hemorrhage that is indicated by red and green arrows. Plaque area indicated by red arrows in Fig 3 shows a bright signal on MPRAGE and T1w, low ADC value (0.32x10^-3mm²/s). The mean ADC value of green arrow area was 0.92x10^-3mm²/s. The mean value of ADC in hemorrhage averaged from the values of six patients was 0.72±0.35x10^-3mm²/sOur sequence was able to yield high resolution ADC maps that could provide clear ROI selection for important plaque components. Our measured ADC values match the values reported in recent in-vivo studies.^1,2 As seen in Fig 3, we observed a difference in ADC values from ROIs selected at two different hemorrhage locations. Some hemorrhage with high water content, such as necrosis may show bright intensity in MPRAGE with low ADC values.⁶ The ADC values of the normal carotid wall compared with the lower ADC values in intraplaque lipid and hemorrhage suggest that ADC measurements may be of substantial value in plaque component discrimination. We believe that this 3D SOS technique can be used to further investigate the ADC in other plaque components.Increased spatial resolution in motion insensitive 3D DW-DE SOS can improve the sensitivity of ADC maps in plaque component identification. The results obtained indicate that an ADC map may be of substantial value in identifying lipid and hemorrhage within overall plaque burden.