Acute nonvariceal upper gastrointestinal (UGI) hemorrhage is a frequent complication associated with significant morbidity and mortality. The most common cause of UGI bleeding is peptic ulcer disease, but the differential diagnosis is diverse and includes tumors; ischemia; gastritis; arteriovenous malformations, such as Dieulafoy lesions; Mallory-Weiss tears; trauma; and iatrogenic causes. Aggressive treatment with early endoscopic hemostasis is essential for a favorable outcome. However, severe bleeding despite conservative medical treatment or endoscopic intervention occurs in 5-10% of patients, requiring surgery or transcatheter arterial embolization. Surgical intervention is usually an expeditious and gratifying endeavor, but it can be associated with high operative mortality rates. Endovascular management using superselective catheterization of the culprit vessel, «sandwich» occlusion, or blind embolization has emerged as an alternative to emergent operative intervention for high-risk patients and is now considered the first-line therapy for massive UGI bleeding refractory to endoscopic treatment. Indeed, many published studies have confirmed the feasibility of this approach and its high technical and clinical success rates, which range from 69 to 100% and from 63 to 97%, respectively, even if the choice of the best embolic agent among coils, cyanaocrylate glue, gelatin sponge, or calibrated particles remains a matter of debate. However, factors influencing clinical outcome, especially predictors of early rebleeding, are poorly understood, and few studies have addressed this issue. This review of the literature will attempt to define the role of embolotherapy for acute nonvariceal UGI hemorrhage that fails to respond to endoscopic hemostasis and to summarize data on factors predicting angiographic and embolization failure.

Cone-beam computed tomography (CBCT) is an imaging technique that provides computed tomographic (CT) images from a rotational scan acquired with a C-arm equipped with a flat panel detector. Utilizing CBCT images during interventional procedures bridges the gap between the world of diagnostic imaging (typically three-dimensional imaging but performed separately from the procedure) and that of interventional radiology (typically two-dimensional imaging). CBCT is capable of providing more information than standard two-dimensional angiography in localizing and/or visualizing liver tumors (“seeing” the tumor) and targeting tumors though precise microcatheter placement in close proximity to the tumors (“reaching” the tumor). It can also be useful in evaluating treatment success at the time of procedure (“assessing” treatment success). CBCT technology is rapidly evolving along with the development of various contrast material injection protocols and multiphasic CBCT techniques. The purpose of this article is to provide a review of the principles of CBCT imaging, including purpose and clinical evidence of the different techniques, and to introduce a decision-making algorithm as a guide for the routine utilization of CBCT during transarterial chemoembolization of liver cancer.

Purpose To evaluate the diagnostic performance of three-dimensional (3D) quantitative enhancement-based and diffusion-weighted volumetric magnetic resonance (MR) imaging assessment of hepatocellular carcinoma (HCC) lesions in determining the extent of pathologic tumor necrosis after transarterial chemoembolization (TACE). Materials and Methods This institutional review board–approved retrospective study included 17 patients with HCC who underwent TACE before surgery. Semiautomatic 3D volumetric segmentation of target lesions was performed at the last MR examination before orthotopic liver transplantation or surgical resection. The amount of necrotic tumor tissue on contrast material–enhanced arterial phase MR images and the amount of diffusion-restricted tumor tissue on apparent diffusion coefficient (ADC) maps were expressed as a percentage of the total tumor volume. Visual assessment of the extent of tumor necrosis and tumor response according to European Association for the Study of the Liver (EASL) criteria was performed. Pathologic tumor necrosis was quantified by using slide-by-slide segmentation. Correlation analysis was performed to evaluate the predictive values of the radiologic techniques. Results At histopathologic examination, the mean percentage of tumor necrosis was 70% (range, 10%–100%). Both 3D quantitative techniques demonstrated a strong correlation with tumor necrosis at pathologic examination (R2 = 0.9657 and R2 = 0.9662 for quantitative EASL and quantitative ADC, respectively) and a strong intermethod agreement (R2 = 0.9585). Both methods showed a significantly lower discrepancy with pathologically measured necrosis (residual standard error [RSE] = 6.38 and 6.33 for quantitative EASL and quantitative ADC, respectively), when compared with non-3D techniques (RSE = 12.18 for visual assessment). Conclusion This radiologic-pathologic correlation study demonstrates the diagnostic accuracy of 3D quantitative MR imaging techniques in identifying pathologically measured tumor necrosis in HCC lesions treated with TACE.

Purpose To evaluate the feasibility of image fusion (IF) of preprocedural arterial-phase computed tomography with intraprocedural fluoroscopy for roadmapping in endovascular repair of complex aortic aneurysms, and to compare this approach versus current roadmapping methods (ie, two-dimensional [2D] and three-dimensional [3D] angiography). Materials and Methods Thirty-seven consecutive patients with complex aortic aneurysms treated with endovascular techniques were retrospectively reviewed; these included aneurysms of digestive and/or renal arteries and pararenal and juxtarenal aortic aneurysms. All interventions were performed with the same angiographic system. According to the availability of different roadmapping software, patients were successively placed into three intraprocedural image guidance groups: (i) 2D angiography (n = 9), (ii) 3D rotational angiography (n = 14), and (iii) IF (n = 14). X-ray exposure (dose–area product [DAP]), injected contrast medium volume, and procedure time were recorded. Results Patient characteristics were similar among groups, with no statistically significant differences (P ≥ .05). There was no statistical difference in endograft deployment success between groups (2D angiography, eight of nine patients [89%]; 3D angiography and IF, 14 of 14 patients each [100%]). The IF group showed significant reduction (P < .0001) in injected contrast medium volume versus other groups (2D, 235 mL ± 145; 3D, 225 mL ± 119; IF, 65 mL ± 28). Mean DAP values showed no significant difference between groups (2D, 1,188 Gy · cm2 ± 1,067; 3D, 984 Gy · cm2 ± 581; IF, 655 Gy · cm2 ± 457; P = .18); nor did procedure times (2D, 233 min ± 123; 3D, 181 min ± 53; IF, 189 min ± 60; P = .59). Conclusions The use of IF-based roadmapping is a feasible technique for endovascular complex aneurysm repair associated with significant reduction of injected contrast agent volume and similar x-ray exposure and procedure time.

Purpose To demonstrate that hepatic tumor volume and enhancement pattern measurements can be obtained in a time efficient and reproducible manner on a voxel-by-voxel basis to provide a true 3D volumetric assessment. Materials and Methods Retrospective evaluation of MRI data obtained from 20 patients recruited for a single-institution prospective study. All patients carried a diagnosis of hepatocellular carcinoma (HCC) and underwent drug-eluting beads transcatheter arterial chemoembolization (DEB-TACE) for the first time. All patients had undergone contrast-enhanced MRI before and after DEB-TACE although poor image quality excluded 3 resulting in a final count of 17 patients. vRECIST and qEASL were measured and segmentation and processing times were recorded. Results Thirty-four scans were analyzed. The time for semi-automatic segmentation was 65±33 seconds with a range of 40–200 seconds. vRECIST and qEASL of each tumor were then computed less than one minute for each. Conclusion Semi-automatic quantitative tumor enhancement (qEASL) and volume (vRECIST) assessment is feasible in a workflow efficient time frame. Clinical correlation is necessary, but vRECIST and qEASL could become part of the assessment of intra-arterial therapy for interventional radiologists.

Purpose To compare currently available non-three-dimensional methods (Response Evaluation Criteria in Solid Tumors [RECIST], European Association for Study of the Liver [EASL], modified RECIST [mRECIST[) with three-dimensional (3D) quantitative methods of the index tumor as early response markers in predicting patient survival after initial transcatheter arterial chemoembolization (TACE). Materials and Methods This was a retrospective single-institution HIPAA-compliant and institutional review board–approved study. From November 2001 to November 2008, 491 consecutive patients underwent intraarterial therapy for liver cancer with either conventional TACE or TACE with drug-eluting beads. A diagnosis of hepatocellular carcinoma (HCC) was made in 290 of these patients. The response of the index tumor on pre- and post-TACE magnetic resonance images was assessed retrospectively in 78 treatment-naïve patients with HCC (63 male; mean age, 63 years ± 11 [standard deviation]). Each response assessment method (RECIST, mRECIST, EASL, and 3D methods of volumetric RECIST [vRECIST] and quantitative EASL [qEASL]) was used to classify patients as responders or nonresponders by following standard guidelines for the uni- and bidimensional measurements and by using the formula for a sphere for the 3D measurements. The Kaplan-Meier method with the log-rank test was performed for each method to evaluate its ability to help predict survival of responders and nonresponders. Uni- and multivariate Cox proportional hazard ratio models were used to identify covariates that had significant association with survival. Results The uni- and bidimensional measurements of RECIST (hazard ratio, 0.6; 95% confidence interval [CI]: 0.3, 1.0; P = .09), mRECIST (hazard ratio, 0.6; 95% CI: 0.6, 1.0; P = .05), and EASL (hazard ratio, 1.1; 95% CI: 0.6, 2.2; P = .75) did not show a significant difference in survival between responders and nonresponders, whereas vRECIST (hazard ratio, 0.6; 95% CI: 0.3, 1.0; P = .04), qEASL (Vol) (hazard ratio, 0.5; 95% CI: 0.3, 0.9; P = .02), and qEASL (%) (hazard ratio, 0.3; 95% CI: 0.15, 0.60; P < .001) did show a significant difference between these groups. Conclusion The 3D-based imaging biomarkers qEASL and vRECIST were tumor response criteria that could be used to predict patient survival early after initial TACE and enabled clear identification of nonresponders.

Transarterial chemoembolization outcomes in patients with HCC may be predicted before procedures by combining clinical patient data and baseline MR imaging with the use of AI and ML techniques.

A deep learning model trained on some labeled data from a certain source domain generally performs poorly on data from different target domains due to domain shifts. Unsupervised domain adaptation methods address this problem by alleviating the domain shift between the labeled source data and the unlabeled target data. In this work, we achieve cross-modality domain adaptation, i.e. between CT and MRI images, via disentangled representations. Compared to learning a one-toone mapping as the state-of-art CycleGAN, our model recovers a manyto-many mapping between domains to capture the complex cross-domain relations. It preserves semantic feature-level information by finding a shared content space instead of a direct pixelwise style transfer. Domain adaptation is achieved in two steps. First, images from each domain are embedded into two spaces, a shared domain-invariant content space and a domain-specific style space. Next, the representation in the content space is extracted to perform a task. We validated our method on a crossmodality liver segmentation task, to train a liver segmentation model on CT images that also performs well on MRI. Our method achieved Dice Similarity Coefficient (DSC) of 0.81, outperforming a CycleGAN-based method of 0.72. Moreover, our model achieved good generalization to joint-domain learning, in which unpaired data from different modalities are jointly learned to improve the segmentation performance on each individual modality. Lastly, under a multi-modal target domain with significant diversity, our approach exhibited the potential for diverse image generation and remained effective with DSC of 0.74 on multi-phasic MRI while the CycleGAN-based method performed poorly with a DSC of only 0.52.