Three-dimensional models of the abdominal vasculature based on helical CT: usefulness in patients with pancreatic neoplasms

When Deep Learning Meets Data Alignment: A Review on Deep Registration Networks (DRNs)

This paper reviews recent deep learning-based registration methods. Registration is the process that computes the transformation that aligns datasets, and the accuracy of the result depends on multiple factors. The most significant factors are the size of input data; the presence of noise, outliers and occlusions; the quality of the extracted features; real-time requirements; and the type of transformation, especially those defined by multiple parameters, such as non-rigid deformations. Deep Registration Networks (DRNs) are those architectures trying to solve the alignment task using a learning algorithm. In this review, we classify these methods according to a proposed framework based on the traditional registration pipeline. This pipeline consists of four steps: target selection, feature extraction, feature matching, and transform computation for the alignment. This new paradigm introduces a higher-level understanding of registration, which makes explicit the challenging problems of traditional approaches. The main contribution of this work is to provide a comprehensive starting point to address registration problems from a learning-based perspective and to understand the new range of possibilities.

Multi-detector Computer Tomography Angiography in the Initial Assessment of Patients Acutely Suspected of Having Intracranial Aneurysm Rupture

Introduction: Multi-detector computer tomography angiography (CTA) provides a fast non-invasive assessment of the cerebral vessels, is readily available in an acute setting and can potentially replace invasive digital subtraction angiography (DSA) forthe diagnosis of intracranial vascular lesions in an emergency setting. We report our experience in the use of emergent cerebral CTA versus DSA in the assessment of patients presenting acutely with symptoms suspicious of brain aneurysm rupture. Materials and Methods: Thirty-seven consecutive patients presenting acutely with clinical suspicion of brain aneurysm rupture were evaluated over a 4-month period from January to April 2008. CTA with peripheral intravenous contrast injection was performed on a 32 slice helical scanner. DSA was performed within 48 hours for all cases when CTA was the initial assessment. Studies were assessed via radiology reports using DSA or surgery as the gold standard. Results: All except for 3 patients had CTA as the initial study. There were 26 cerebral aneurysms detected by CTA in these 37 patients, with 9 negative studies. There were 2 patients with arteriovenous malformations (AVM), 1 with AV fistula (AVF), 1 tumoral bleed, 2 vertebral dissections, and 1 missed sagittal sinus thrombosis (CVT) on CTA. Based solely on CTA assessment, 3 patients had direct surgical clipping of the aneurysm, while 4 proceeded to direct endovascular coiling. Conclusion: Emergent CTA is a non-invasive, reliable and viable alternative to emergent DSA for the assessment of the cerebral vessels in the acute assessment of patients presenting with symptoms suspicious of brain aneurysm rupture. Where positive, it can serve as a guide to therapeutic decisions. Review of CTA source data is essential, especially for small lesions and for post-clipping assessment. Key words: Digital subtraction angiography, Intracranial aneurysms, Multi-detector computer tomography angiography

Pancreatic adenocarcinoma: analysis of the effect of various concentrations of contrast material

PURPOSE

The aim of this study was to compare the efficacy of two contrast materials with moderate and high iodine concentrations for the depiction of pancreatic adenocarcinoma.

MATERIALS AND METHODS

A series of 107 patients with histologically proven pancreatic adenocarcinoma underwent helical computed tomography. A fixed dose of 100 ml of iopamidol 300 (mg I/ml) was administered to 50 patients (group A) and iopamidol 370 (mg I/ml) to 57 patients (group B) at the same injection rate (3 ml/s). Unenhanced helical scans and contrast-enhanced scans for three phases (30, 70, and 300 s after starting the infusion of contrast material) were obtained. We evaluated enhancement of the aorta, portal vein, hepatic parenchyma, pancreatic parenchyma, and pancreatic adenocarcinoma during each phase.

RESULTS

During all phases, both aortic and pancreatic enhancement were significantly greater in group B than in group A (P<0.01). Enhancement of the portal vein and hepatic parenchyma was significantly greater at 70 and 300 s in group B than in group A (both P<0.01). Tumor-to-pancreas contrast was significantly greater in group B than in group A at both 30 s (P<0.01) and 70 s (P<0.05).

CONCLUSION

Administration of contrast material with a high iodine concentration is more effective for depicting pancreatic adenocarcinomas.

3-Dimensional computerized tomography in follow-up of patients with urinary diversion

INTRODUCTION

The aim of our study was to evaluate the usefulness of 3-dimensional computerized tomography (3D-CT) in routine follow-up of patients who had undergone radical cystectomy and different kinds of urinary diversions and compare it with conventional CT.

PATIENTS AND METHODS

Nineteen patients (18 men, 1 woman) who had undergone radical cystectomy and different kinds of urinary diversions with diagnosis of invasive bladder cancer were enrolled into the study. The mean age of the patients was 55.5 (46-69) years. For all patients, conventional CT was performed, followed by 3D reconstruction of these images.

RESULTS

Orthotopic ileal neobladder (Abol-Enein and Ghoneim procedure) was performed in 12, rectosigmoid pouch (Mainz pouch II) in 2 and ileal conduit in 5 patients. There were no pathological findings visible either on conventional CT or on 3D-CT, but the shape, configuration, and the relationships of the pouch with ureters, urethra and other adjacent organs were much better visualized on 3D-CT.

CONCLUSION

3D-CT did not have any advantages over conventional CT for showing pathological findings that were oncological origin, but as the anatomy of the lower urinary tract can be perfectly demonstrated, we think that it may be used for planning of a re-operation after urinary diversion and may help the urologist who has less experience with radiological studies to evaluate pouch configuration and indentations to the adjacent organs better.

[Imaging of pancreatic adenocarcinoma]

Pancreatic cancer remains the fourth most common cause of cancer death. Surgery remains the only option for cure. Accurate diagnosis and staging are essential for appropriate management of patients with pancreatic cancer. This paper reviews the state of the art for imaging modalities in the diagnosis and staging of pancreatic adenocarcinoma. The crucial role of CT has increased with the new generation of multidetector CT.

Vascular involvement in pancreatic neoplasm: a comparison between spiral CT and DSA

The principal criterion for resectability of pancreatic carcinoma is the assessment of vascular involvement. In a prospective evaluation the ability of Spiral CT Angiography (CTA) to detect vascular involvement in 50 patients with pancreatic carcinoma, was proved; DSA was performed later in all patients. In 20 patients, without vascular involvement, a complete concordance was obtained. Of 30 patients with vascular involvement, there was complete concordance between CTA and angiography in 22 patients and discordance in 8 patients. CTA was superior in 2 cases with periadventitial infiltration and in 5 patients with splenoportal confluence thrombosis. DSA was superior in 1 case with infiltration of the superior mesenteric vein. After surgical evaluation, sensitivity of CTA and DSA was 97% and 77%, respectively, and the negative predictive values were 95% and 74%. As compared to DSA, CTA is more rapid and less invasive and can be considered the modality of choice for preoperative work-up of pancreatic neoplasm.

Helical CT with CT angiography in assessing periampullary neoplasms: identification of vascular invasion

PURPOSE

To determine the accuracy of helical computed tomography (CT) with CT angiography in identifying vascular invasion by periampullary neoplasms and to assess the added value of CT angiography.

MATERIALS AND METHODS

Sixty-nine patients suspected of having periampullary neoplasms were examined. Images from dual phase helical CT with CT angiography were compared with surgical findings in 36 patients. Arterial and venous invasion were assessed separately. Accuracy, positive predictive value (PPV), and negative predictive value (NPV) were determined for CT alone and for CT supplemented with CT angiography.

RESULTS

The accuracy, PPV, and NPV of helical CT with CT angiography in identifying venous invasion was 92% (33 of 36 patients), 86% (12 of 14 patients), and 95% (21 of 22 patients), respectively. When transverse CT images alone were analyzed, accuracy decreased to 69% (25 of 36 patients) (P =.005); PPV and NPV were 63% (five of eight patients) and 71% (20 of 28 patients), respectively. When identifying arterial invasion, the accuracy of CT with CT angiography and of CT alone was 86% (31 of 36 patients). PPV and NPV also were identical at 71% (five of seven patients) and 90% (26 of 29 patients), respectively.

CONCLUSION

CT angiography significantly increases the ability to identify venous invasion when compared with CT alone but does not improve detection of arterial invasion.

Evaluation of extrahepatic portal hypertension and surgical portal systemic shunts by intravenous CT portography

The purpose of this study was to evaluate the splenoportal (SP) axis and patency of portal systemic shunts in extrahepatic portal venous obstruction (EHO) by intravenous CT portography (CTP). Fifty-five patients of preshunt EHO, and 21 patients of postshunt (surgical portal systemic shunts) EHO, were subjected to intravenous CTP on a subsecond helical CT scanner. Thin, axial sections and three dimensional (3-D) reconstructions, including maximum intensity projection (MIP) and shaded surface display (SSD), were obtained. The findings were correlated with Color Doppler Flow Imaging (CDFI). In the EHO preshunt group, the site of the block demonstrated on CTP correlated with CDFI in 53 of 55 cases. In the postshunt group, shunt patency evaluation by CTP was in agreement with CDFI in 19 of 21 patients (8 patent; 13 blocked). Two patients in each group could not be evaluated on CDFI, while CTP could provide the appropriate information. There was no false positivity or negativity with CTP in patients evaluated on both modalities. The etiology of EHO, the global view of collaterals, and the 3-D anatomy of SP axis could be well depicted and was well accepted by the surgeons. We conclude that CTP is an impressive new technique that can effectively evaluate pre- and postshunt cases of EHO.

Local extension of pancreatic carcinoma: assessment with thin-section helical CT versus with breath-hold fast MR imaging--ROC analysis

PURPOSE

To compare contrast material-enhanced thin-section helical CT with breath-hold contrast-enhanced MR imaging for sensitivity in the detection of pancreatic adenocarcinoma and for accuracy in local tumor staging.

MATERIALS AND METHODS

Fifty-seven patients (37 men, 20 women aged 42-28 years) suspected of having pancreatic adenocarcinoma were examined. The final diagnosis was confirmed at surgery to be pancreatic cancer in 31 patients; the other 26 patients were deemed not to have pancreatic cancer. All patients underwent both CT and MR imaging (turbo spin-echo and fast low-angle shot) studies. Image quality and pancreatic enhancement were subjectively evaluated. All CT scans and MR images were assessed by two independent observers by using a five-point scale for the detection of tumor and of invasion into the peripancreatic tissue, portal vein, and/or peripancreatic artery. Receiver operating characteristic curves for CT and MR imaging were analyzed.

RESULTS

At visual analysis, pancreatic enhancement at CT and at MR imaging was comparable, but depiction of vessels was superior at helical CT. Detectability of tumor was comparable. Helical CT was significantly superior to MR imaging in diagnostic imaging of invasion into the peripancreatic tissue, portal vein, and/or peripancreatic artery (P < .01).

CONCLUSION

Thin-section dynamic CT is more sensitive than MR imaging for detection of peripancreatic and vascular invasion in patients with pancreatic cancer.

CT angiography of the visceral vasculature

CT angiography (CTA) is a minimally invasive technique that has proven to be clinically useful in evaluating the vasculature of the abdominal viscera. In many instances, the diagnostic information obtained from abdominal CTA is sufficient to avoid the expense and morbidity of conventional angiography. This article reviews the indications, technique, and pitfalls of abdominal CTA with specific emphasis on disorders of the hepatic, splenic, and superior mesenteric arteries and the portal venous system.

Three-dimensional spiral CT angiography in pancreatic surgical planning using non-tailored protocols: comparison with conventional angiography

The aim of this study was to investigate three-dimensional spiral computed tomography (3DCT) as an adjunct to routine pancreatic CT scanning, with particular regard to the identification of surgically important hepatic arterial anomalies, correlated with conventional visceral angiography. 32 patients underwent spiral CT scans prior to pancreatic surgery using established protocols. Oral contrast medium was used throughout. 150 ml of intravenous contrast medium was given at 3 ml s-1 with a 24 s spiral CT sequence starting 35 s after the start of infusion. Two protocols were employed, both with a pitch of 1:3 mm table feed/collimation (n = 17) and 5 mm table feed/collimation (n = 15). Overlapping (1 mm minimum) axial reformats were reconstructed. 3DCT shaded-surface displays of the visceral arteries were assessed for visceral arterial anomalies. Visceral angiography (n = 23) was independently correlated. Satisfactory 3D angiograms were performed in all but one patient, in whom the coeliac axis was missed. (i) 3 mm protocol: 3DCT (n = 17) showed three anomalous right hepatic arteries (ARHA), one trifurcation anomaly and one splenic artery with an aortic origin. Angiography (n = 11) confirmed these findings, although one patient with an ARHA did not have angiography. A left gastric arterial supply to the left liver was not detected. (ii) 5 mm protocol: 3DCT (n = 15) showed two cases of ARHA. While confirming these findings, angiography (n = 12) showed a third case of ARHA, in which the coeliac and superior mesenteric artery had very close origins. A left gastric supply to the left liver was also missed. It is concluded that satisfactory 3DCT is possible without changing existing scanning protocols, although narrow sections are required for the confident assessment of right hepatic arterial anomalies, and any left hepatic supply via the left gastric artery was poorly assessed in this series.

Focal nodular hyperplasia of the liver: assessment of hemodynamic and angioarchitectural patterns with gadolinium chelate-enhanced 3D spoiled gradient-recalled MRI and maximum intensity projection reformatted images

PURPOSE

Our goal was to determine the relative merits of gadolinium chelate-enhanced 3D spoiled gradient-recalled (GRE) MRI versus maximum intensity projection (MIP) reformatted images in assessing the morphologic, hemodynamic, and angioarchitectural patterns of focal nodular hyperplasia (FNH) of the liver.

METHOD

Ten consecutive patients with 10 FNHs had prospectively gadolinium chelate-enhanced 3D spoiled GRE MRI (TR/TE/FA = 10.1/1.9/30) of the liver at 1.5 T. Gadolinium chelate-enhanced 3D spoiled GRE source images and MIP reformatted images were separately analyzed with respect to morphologic and hemodynamic features and angioarchitectural patterns by two independent readers.

RESULTS

Gadolinium chelate-enhanced 3D spoiled GRE source images and MIP reformatted images showed the most intense degrees of enhancement of FNH during the arterial phase of hepatic parenchymal enhancement in all cases. Gadolinium chelate-enhanced 3D spoiled GRE source images were superior to MIP reformatted images for the assessment of morphologic features of FNH (p < 0.02). MIP reformatted images were superior to the corresponding source images for showing the main branches of the hepatic artery, an arterial branch going to the FNH, and a small artery within the FNH radiating to peripheral areas (p < 0.05). There was excellent agreement between the two observers for analysis of the MIP reformatted images (p < 0.05).

CONCLUSION

The combination of gadolinium chelate-enhanced 3D spoiled GRE source images and MIP reformatted images allows the analysis of morphologic, hemodynamic, and angioarchitectural patterns of FNH of the liver. Further study and comparison with currently applied strategies will determine the value of these two techniques for diagnosing FNH of the liver.

CT angiography of the visceral vessels

CT angiography (CTA) is a promising new technique for vascular imaging. This review focuses first on the technique necessary for successful scanning of the visceral vessels. As in many new modalities, there are different protocols for scanning and rendering of images. The relative strengths and weaknesses of these different approaches are discussed. A discussion of the applications of CTA to depict normal and abnormal anatomy of the visceral vessels follows. These applications include celiac stenosis, splenic artery aneurysms, evaluation for hepatic arterial anatomy before liver transplantation, visceral arterial anatomy in pancreaticoduodenal surgery, the superior mesenteric artery in intestinal ischemia, vascular encasement in patients with pancreatic neoplasms, and, finally, the hepatic vessels before hepatic tumor resection.

Three-dimensional helical CT of intrahepatic venous structures: comparison of three rendering techniques

OBJECTIVE

To define the advantages and disadvantages of various rendering techniques to obtain three-dimensional (3D) displays of intrahepatic venous structures with helical CT data.

MATERIALS AND METHODS

After rapid preprocessing segmentation of the liver, helical CT data (8 mm slice thickness overlapped every 4 mm) from 10 patients were reconstructed using maximum intensity projection (MIP), volume rendering, and surface rendering algorithms. Three-dimensional imaging was evaluated blindly and independently by three observers for presence of artifacts and overall quality.

RESULTS

Three-dimensional displays showed the hepatic veins and fifth order portal branches with the volume and MIP rendering techniques. Best overall quality in the 3D representation of the liver was achieved with the MIP technique (p < 0.05). Small details in venous anatomy and portal involvement by tumor were better imaged with the MIP technique. "Stair-step" artifacts markedly degraded the 3D displays obtained with the surface rendering technique, making it inappropriate for imaging the intrahepatic venous structures.

CONCLUSION

Maximum intensity projection appears to be an adequate technique to perform 3D imaging of intrahepatic venous structures with helical CT data when 8 mm slice thicknesses overlapping every 4 mm are used. However, optimization of imaging protocols needs to be done and compared in a larger series.

The current state of the art in three dimensional oncologic imaging: an overview

PURPOSE

To provide an overview of the methods and clinical applications of three dimensional (3D) medical imaging in the oncologic patient.

METHODS AND MATERIALS

We briefly outline the techniques currently used to create 3D medical images with an emphasis on their strengths and shortcomings as they relate to oncologic imaging and radiation therapy planning. We then discuss some of the most important and promising oncologic applications of 3D imaging and suggest likely future directions in this rapidly developing field.

RESULTS

Since the first application of 3D techniques to medical data over a decade ago, 3D medical images have evolved from relatively crude representations of musculoskeletal abnormalities to detailed and accurate representations of a variety of soft tissue, vascular, and oncologic pathology. The rapid development of both computer hardware and software coupled with the application of 3D techniques to a variety of imaging modalities have expanded the clinical applications of this technology dramatically.

CONCLUSIONS

3D medical images are clinically practical tools for oncologic evaluation and effective radiation therapy planning.