Comparison of utility of deep learning reconstruction on 3D MRCPs obtained with three different k-space data acquisitions in patients with IPMN.

OBJECTIVE: To compare the utility of deep learning reconstruction (DLR) for improving acquisition time, image quality, and intraductal papillary mucinous neoplasm (IPMN) evaluation for 3D MRCP obtained with parallel imaging (PI), multiple k-space data acquisition for each repetition time (TR) technique (Fast 3D mode multiple: Fast 3Dm) and compressed sensing (CS) with PI. MATERIALS AND METHODS: A total of 32 IPMN patients who had undergone 3D MRCPs obtained with PI, Fast 3Dm, and CS with PI and reconstructed with and without DLR were retrospectively included in this study. Acquisition time, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) obtained with all protocols were compared using Tukey's HSD test. Results of endoscopic ultrasound, ERCP, surgery, or pathological examination were determined as standard reference, and distribution classifications were compared among all 3D MRCP protocols by McNemar's test. RESULTS: Acquisition times of Fast 3Dm and CS with PI with and without DLR were significantly shorter than those of PI with and without DLR (p < 0.05). Each MRCP sequence with DLR showed significantly higher SNRs and CNRs than those without DLR (p < 0.05). IPMN distribution accuracy of PI with and without DLR and Fast 3Dm with DLR was significantly higher than that of Fast 3Dm without DLR and CS with PI without DLR (p < 0.05). CONCLUSION: DLR is useful for improving image quality and IPMN evaluation capability on 3D MRCP obtained with PI, Fast 3Dm, or CS with PI. Moreover, Fast 3Dm and CS with PI may play as substitution to PI for MRCP in patients with IPMN. KEY POINTS: • Mean examination times of multiple k-space data acquisitions for each TR and compressed sensing with parallel imaging were significantly shorter than that of parallel imaging (p < 0.0001). • When comparing image quality of 3D MRCPs with and without deep learning reconstruction, deep learning reconstruction significantly improved signal-to-noise ratio and contrast-to-noise ratio (p < 0.05). • IPMN distribution accuracies of parallel imaging with and without deep learning reconstruction (with vs. without: 88.0% vs. 88.0%) and multiple k-space data acquisitions for each TR with deep learning reconstruction (86.0%) were significantly higher than those of others (p < 0.05).

Preoperative CT Findings for Predicting Acute Exacerbation of Interstitial Pneumonia After Lung Cancer Surgery: A Multicenter Case-Control Study.

BACKGROUND. Acute exacerbation (AE) is a life-threatening complication of inter-stitial pneumonia (IP). Thoracic surgery may trigger AE. OBJECTIVE. The purpose of this study is to explore the role of preoperative CT findings in predicting postoperative AE in patients with IP and lung cancer. METHODS. This retrospective case-control study included patients from 22 institutions who had IP and underwent thoracic surgery for lung cancer. AE was diagnosed on the basis of symptoms and imaging findings noted within 30 days after surgery and the absence of alternate causes. For each patient with AE, two control patients without AE were identified. After exclusions, the study included 92 patients (78 men and 14 women; 31 with AE [the AE group] and 61 without AE [the no-AE group]; mean age, 72 years). Two radiologists independently reviewed preoperative thin-slice CT examinations for pulmonary findings and resolved differences by consensus. The AE and no-AE groups were compared using the Fisher exact and Mann-Whitney U tests. Multivariable logistic regression was performed. Interreader agreement was assessed by kappa coefficients. RESULTS. A total of 94% of patients in the AE group underwent segmentectomy or other surgery that was more extensive than wedge resection versus 75% in the no-AE group (p = .046). The usual IP pattern was present in 58% of the AE group versus 74% of the no-AE group (p = .16). According to subjective visual scoring, the mean (± SD) ground-glass opacity (GGO) extent was 6.3 ± 5.4 in the AE group versus 3.9 ± 3.8 in the no-AE group (p = .03), and the mean consolidation extent was 0.5 ± 1.2 in the AE group versus 0.1 ± 0.3 in the no-AE group (p = .009). Mean pulmonary trunk diameter was 28 ± 4 mm in the AE group versus 26 ± 3 mm in the no-AE group (p = .02). In a model of CT features only, independent predictors of AE (p < .05) were GGO extent (odds ratio [OR], 2.8), consolidation extent (OR, 9.4), and pulmonary trunk diameter (OR, 4.2); this model achieved an AUC of 0.75, a PPV of 71%, and an NPV of 77% for AE. When CT and clinical variables were combined, undergoing segmentectomy or more extensive surgery also independently predicted AE (OR, 8.2; p = .02). CONCLUSION. The presence of GGO, consolidation, and pulmonary trunk enlargement on preoperative CT predicts AE in patients with IP who are undergoing lung cancer surgery. CLINICAL IMPACT. Patients with IP and lung cancer should be carefully managed when predictive CT features are present. Wedge resection, if possible, may help reduce the risk of AE in these patients. TRIAL REGISTRATION. University Hospital Medical Information Clinical Trial Registry UMIN000029661.

Urgent intraoperative endovascular stent placement to resolve acute hepatic or portal venous obstruction during liver surgery: a case series.

BACKGROUND: Acute obstruction of the hepatic vein (HV) or the portal vein (PV), particularly when it occurs during liver surgery, is potentially fatal unless repaired swiftly. As surgical interventions for this problem are technically demanding and potentially unsuccessful, other treatment options are needed. CASE PRESENTATION: We report two cases of acute, surgically uncorrectable HV or PV obstruction during liver resection or living donor liver transplantation (LDLT), which was successfully treated with urgent intraoperative placement of endovascular stents using interventional radiology (IVR). In Case 1, a patient with colonic liver metastases underwent a non-anatomic partial hepatectomy of the segments 4 and 8 with middle hepatic vein (MHV) resection. Additionally, the patient underwent an extended right posterior sectionectomy with right hepatic vein (RHV) resection for tumors involving RHV. Reconstruction of the MHV was needed to avoid HV congestion of the anterior section of the liver. The MHV was firstly reconstructed by an end-to-end anastomosis between the MHV and RHV resected stumps. However, the reconstruction failed to retain the HV outflow and the anterior section became congested. Serial trials of surgical revisions including re-anastomosis, vein graft interposition and vein graft patch-plasty on the anastomotic wall failed to recover the HV outflow. In Case 2, a pediatric patient with biliary atresia underwent an LDLT and developed an intractable PV obstruction during surgery. Re-anastomosis with vein graft interposition failed to restore the PV flow and elongated warm ischemic time became critical. In both cases, the misalignment in HV or PV reconstruction was likely to have caused flow obstruction, and various types of surgical interventions failed to recover the venous flow. In both cases, an urgent IVR-directed placement of self-expandable metallic stents (SEMS) restored the HV or PV perfusion quickly and effectively, and saved the patients from developing critical conditions. Furthermore, in Cases 1 and 2, the SEMS placed were patent for a sufficient period of time (32 and 44 months, respectively). CONCLUSIONS: The IVR-directed, urgent, intraoperative endovascular stenting is a safe and efficient treatment tool that serves to resolve the potentially fatal acute HV or PV obstruction that occurs in the middle of liver surgery.

Biochemical and Clinical Predictive Approach and Time Point Analysis of Hepatobiliary Phase Liver Enhancement on Gd-EOB-DTPA-enhanced MR Images: A Multicenter Study.

Purpose To identify biochemical factors associated with liver enhancement over time on gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance (MR) images and predict the optimal time point of the hepatobiliary phase in various clinical settings. Materials and Methods This study was approved by the institutional review boards, and written informed consent was obtained from the 1903 patients enrolled. Simple and multiple logistic regression analyses were performed to investigate the biochemical factors associated with liver-to-spleen contrast (LSC) of at least 1.5 in the hepatobiliary phase. Changes in LSC and lesion-to-liver contrast (LLC) of lesions over time (at 5, 10, 15, and 20 minutes) were investigated with a linear mixed-effects model in patients and lesions. For LSC, the optimal cutoff value was determined with receiver operating characteristic analysis of the most significant variable. Differences in LSC and LLC were analyzed in various clinical settings. Results Ultimately, 1870 patients were evaluated, as 33 were excluded according to study criteria. Prothrombin (PT) activity, total bilirubin level (P = .020), and total cholesterol level (P = .005) were significantly associated with LSC of at least 1.5 at 20 minutes, and PT activity was identified as the most significant factor (odds ratio, 1.271; 95% confidence interval: 1.109, 1.455; P = .001). LSC of at least 1.5 at 20 minutes with PT activity of at least 86.9% and less than 86.9% occurred in 555 of 626 patients (88.6%) and 388 of 575 patients (67.5%), respectively. Satisfactory liver enhancement at 20 minutes was significantly more likely to be achieved by patients with hepatitis B virus than by those with hepatitis C virus (P < .001) and by patients with metastasis than by those with hepatocellular carcinoma (P < .001). No significant difference in LLC was observed in patients examined at 1.5 and 3.0 T (P = .133). Conclusion Hepatic enhancement is significantly associated with PT activity, total bilirubin level, and total cholesterol level. PT activity of at least 86.9% could be used to shorten examination times at Gd-EOB-DTPA-enhanced MR imaging. © RSNA, 2016 Online supplemental material is available for this article.

Hybrid repair of an aortocaval fistula and inferior vena cava external compression caused by an inflammatory aortoiliac aneurysm: a case study.

INTRODUCTION: We report a case of aortocaval fistula successfully treated by hybrid operation. CASE DESCRIPTION: A 73-year-old female suffering from malignant lymphoma and painful leg edema was transferred to our institution. Computed tomography revealed an aortoiliac aneurysm. The inferior vena cava was compressed by displacement of the abdominal aortic aneurysm. The bilateral internal iliac and ovarian veins were markedly dilated. Diagnosis was an aortoiliac aneurysm with aortocaval fistula. The treatment options were open surgery or an intervention with bypass surgery. Because of narrow iliac access for a bifurcated stent graft, aorto-uni stentgraft treatment followed by bypass surgery was finally decided. Following stent graft insertion and iliofemoral artery bypass, the aneurysms and fistula were successfully excluded without endoleaks. To treat the inferior vena cava compression, the kissing technique was used to place bare metallic stents across the bilateral common iliac veins and inferior vena cava, which improved the clinical symptoms. DISCUSSION AND EVALUATION: In this aortocaval fistula caused by AAA, a minimally invasive treatment of stentgraft and bypass surgery with venous flow recovery was chosen as a hybrid treatment. Intravascular intervention was the most suitable in this situation. Bare stent placement for venous occlusion was also effective for revascularization of vena cava flow. CONCLUSION: Recent advances in endovascular devices, including stent grafts and bare metallic stents, will be helpful for effective noninvasive treatment for aortocaval fistula circulation.

[Navigation surgery for pancreatic disease using multislice computed tomography].

In navigation surgery, preoperatively acquired image data are used so that surgical instruments can be guided inside the body while their location is displayed on a computer monitor. It is used in cranial nerve and spinal surgery. In the field of abdominal surgery, however, surgical manipulations in the target area cause major changes in the displayed images compared with those obtained preoperatively, and therefore, with the exception of certain organs, navigation surgery is difficult to apply. In general, this type of surgery aims to use intraoperative image information to improve surgical precision, carry out the preoperative plan accurately, and avoid dangerous areas. Three-dimensional images of the vascular architecture obtained with multislice computed tomography (MS-CT) make it possible to visualize arteries, the portal vein, bile duct, and even the pancreatic duct from any angle, which cannot be done with conventional angiography. Accurate positional relationships in the affected region can be determined preoperatively by manipulating multiplanar reconstruction images at a work station. MS-CT is extremely useful in navigation for safe performance of all types of pancreatectomy.