Literature Review on DWI?
DIFFUSION ACQUISITION TECHNIQUE FOR LIVER DWI

sequences of the liver typically include at least two b-values: a low b-value (range of 0–100 s/mm2 ) and a high b-value (range of 400–800 s/mm2 ). Low b-values (50–100 s/mm2 ) can suppress intravascular signal, producing black-blood T2-weighted images with good signal-to-noise ratio (SNR) and tissue contrast, enhancing lesion detectability particularly in small lesions that are close to vessels. High b-values have a lower SNR because of signal loss from diffusion, as well as artifacts, but provide better lesion characterization by enhancing differences in signal intensity between liver parenchyma and lesions. The most widely used strategy for DWI is echoplanar imaging (EPI), which allows acquisition of a full slice in a single shot. A typical protocol involves the use of fat-saturated single-shot diffusion-weighted EPI played in an interleaved multislice fashion to allow volume coverage. However, the EPI readout is also subject to “ghosting” and susceptibility artifacts (Le Bihan et al. 2006). Turbo spin-echo and steadystate free-precession imaging (Lu et al. 2012) may offer reduced susceptibility artifacts compared with EPI readouts. However, this usually leads to increased echo trains and to higher-power deposition of the RF pulse train. Periodically rotated overlapping parallel lines with enhanced reconstruction (Deng et al. 2006) is a modified segmented EPI technique whereby successive segments are acquired in a radial fashion. Compared with segmented EPI, this technique may be more immune to motion artifacts, with options for robust motion correction (Deng et al. 2006). Physiologic motion is inherent to any liver imaging protocol, with breathing and cardiac motion resulting in subject- and acquisition-dependent imaging artifacts. Single-shot EPI is robust to motion because its acquisition time is faster than physiologic processes. However, in a typical DWI acquisition with multiple b-values and signal averaging, residual fluctuations exist between successively acquired EPI images. Breath-hold, freebreathing, or respiratory-triggering technique may be used. Breath-hold techniques result in the shortest image acquisition time, with a limitation in the number of bvalues that can be used. Free-breathing protocols typically result in blurred images. Respiratory-triggering acquisition (with the use of navigator echoes or bellows) may improve DWI data quality (Dyvorne et al. 2013) at the cost of increased imaging time. Cardiac motion artifacts usually appear in the left liver lobe as signal loss at a high b-value as a result of strong dephasing of the coherently moving spins under the influence of diffusion gradients. This artifact can be overcome by performing DWI acquisitions at diastole by using an electrocardiogram or pulse trigger (Mürtz et al. 2002), but this approach results in significantly increased scan time. Another promising approach to mitigate signal loss caused by cardiac motion is to use motion-compensated diffusion gradients (Ozaki et al. 2013) to cancel the dephasing of coherently moving tissues while maintaining diffusion weighting.

ASSESSMENT OF HCC RESPONSE TO INTRAARTERIAL THERAPIES BY DWI

Assessment of the efficacy of liver mets. following SIRT is essential for therapeutic decision making, such as whether to repeat, interrupt, or completely terminate therapy. The visual assessment of DWI, which includes images at higher b-values (≥ 500 s/mm2 ), may aid in distinguishing the different components of liver mets. (ie, viable and necrotic components) following SIRT. As a general observation, liver mets. tissues that have become necrotic (with liquefaction or coagulation necrosis) secondary to SIRT typically show lower signal intensity on higher b-value images compared with viable tissues. However, on diffusion images, the signal intensities observed are dependent on not only the water proton diffusion but also the T2 relaxation time of the tissue, which is a possible confounding factor (Goshima et al. 2008). As a result, diffusion images must be interpreted concurrently with the ADC measurements to prevent misinterpretation. The diagnostic performance of ADC quantification for the viable and necrotic components of liver mets. following SIRT was also reported in previous studies (Yuan et al. 2014). Following treatment, the ADC values of the necrotic tumor tissue were greater than those of the viable tumor tissue. After treatment, the tumor ADC value had a strong correlation with the degree of tumor necrosis on pathologic examination (Yuan et al. 2014). Based on a threshold mean ADC value of 1.84 × 10 -3 mm2 /s, receiver operating characteristic analysis showed high sensitivity and specificity values in the identification of necrotic tumor tissues (92.3% and 100%, respectively) following chemoembolization in patients with HCCs (Yuan et al. 2014). ADC has also been used to early evaluate and predict response to intraarterial therapies (Kamel et al. 2006). An increase in ADC values has been reported following radioembolization (Kamel et al. 2007) and chemoembolization (Yuan et al. 2010) in the early posttreatment period (a few days to as long as 2 week), with measurable differences before versus after treatment (Sahin et al. 2012), and treatment effect has been noted at 1–3 months after treatment. Based on a threshold change in ADC value of 16.2%, receiver operating characteristic analysis showed higher sensitivity and specificity values for percentage change in ADC than for absolute change in ADC in the identification of responding lesions after chemoembolization (Kamel et al. 2007) Several researchers have also investigated the role of the pretreatment ADC value in predicting the response to chemoembolization, with discordant results (42,49). The study of Yuan et al (Kamel et al. 2007) showed that nonresponding lesions had significantly higher pretreatment mean ADCs than responding lesions (1.726 ×10 -3 mm2 /s ± 0.323 vs 1.294 × 10 -3 mm2 /s ± 0.185). A high pretreatment mean ADC value was predictive of poor liver mets. response to chemoembolization (Kamel et al. 2007) Mannelli et al (Mannelli et al. 2013) reported that liver mets. with poor and incomplete response to chemoembolization (≤ 50% necrosis on post-chemoembolization MR imaging) had significantly lower pretreatment ADCs and lower SIRT ADCs compared with liver mets. with good/complete response (≥ 50% necrosis). This discordance may be related to the nature of tumors with or without necrotic tissue before treatment (Yuan et al. 2013). Recent publications have also investigated ADCs to predict progression-free survival and/or overall survival in patients with liver mets. (Dong et al. 2012). Pretreatment ADC values, as well as changes in ADC values after treatment, may provide useful information for predicting survival for patients with unresectable liver mets. (Dong et al. 2012) showed that there were significant linear-regression relations between survival time and SIRT lesion ADC values and changes in ADC values after chemoembolization. A log-rank test showed that SIRT ADC values and changes in ADC values after chemoembolization significantly influenced the overall cumulative survival (Dong et al. 2012). Kokabi et al., (2014) reported that a 30% increase in ADC value at 30 days after SIRT is a reproducible early imaging response biomarker that predicts tumor response and prolonged survival following SIRT in infiltrative HCC with portal vein thrombosis. Vandecaveye et al., (2013) showed that 1-month response determined with ADC is an independent predictor of outcome for liver mets. treated with chemoembolization. In summary, DWI may aid in distinguishing viable tumor from necrotic tissue after SURT. After treatment, tumor ADC values have shown strong correlation with the degree of tumor necrosis on pathologic examination. ADC values may also provide useful information for predicting survival for patients with liver mets.