Deep neural network-based segmentation of normal and abnormal pancreas on abdominal CT: evaluation of global and local accuracies.

PURPOSE: Delay in diagnosis can contribute to poor outcomes in pancreatic ductal adenocarcinoma (PDAC), and new tools for early detection are required. Recent application of artificial intelligence to cancer imaging has demonstrated great potential in detecting subtle early lesions. The aim of the study was to evaluate global and local accuracies of deep neural network (DNN) segmentation of normal and abnormal pancreas with pancreatic mass. METHODS: Our previously developed and reported residual deep supervision network for segmentation of PDAC was applied to segment pancreas using CT images of potential renal donors (normal pancreas) and patients with suspected PDAC (abnormal pancreas). Accuracy of DNN pancreas segmentation was assessed using DICE simulation coefficient (DSC), average symmetric surface distance (ASSD), and Hausdorff distance 95% percentile (HD95) as compared to manual segmentation. Furthermore, two radiologists semi-quantitatively assessed local accuracies and estimated volume of correctly segmented pancreas. RESULTS: Forty-two normal and 49 abnormal CTs were assessed. Average DSC was 87.4 ± 3.1% and 85.5 ± 3.2%, ASSD 0.97 ± 0.30 and 1.34 ± 0.65, HD95 4.28 ± 2.36 and 6.31 ± 6.31 for normal and abnormal pancreas, respectively. Semi-quantitatively, ≥95% of pancreas volume was correctly segmented in 95.2% and 53.1% of normal and abnormal pancreas by both radiologists, and 97.6% and 75.5% by at least one radiologist. Most common segmentation errors were made on pancreatic and duodenal borders in both groups, and related to pancreatic tumor including duct dilatation, atrophy, tumor infiltration and collateral vessels. CONCLUSION: Pancreas DNN segmentation is accurate in a majority of cases, however, minor manual editing may be necessary; particularly in abnormal pancreas.

Extramedullary Hematopoiesis: A Forgotten Diagnosis and a Great Mimicker of Malignancy.

ABSTRACT: Radiology errors have been reported in up to 30% of cases when patients have abnormal imaging findings. Although more than half of errors are failures to detect critical findings, over 40% of errors are when findings are recognized but the correct diagnosis or interpretation is not made. One common source of error is when imaging findings from one process simulate imaging findings from another process but the correct diagnosis is not made. This can result in additional imaging studies, unnecessary biopsies, or surgery. Extramedullary hematopoiesis is one of those uncommon disease processes that can produce many imaging findings that may lead to misdiagnosis. The objective of this article is to review the common and uncommon imaging features of extramedullary hematopoiesis while presenting a series of interesting relevant illustrative cases with emphasis on CT.

Classification of pancreatic cystic neoplasms using radiomic feature analysis is equivalent to an experienced academic radiologist: a step toward computer-augmented diagnostics for radiologists.

PURPOSE: A wide array of benign and malignant lesions of the pancreas can be cystic and these cystic lesions can have overlapping imaging appearances. The purpose of this study is to compare the diagnostic accuracy of a radiomics-based pancreatic cyst classifier to an experienced academic radiologist. METHODS: In this IRB-approved retrospective single-institution study, patients with surgically resected pancreatic cysts who underwent preoperative abdominal CT from 2003 to 2016 were identified. Pancreatic cyst(s) and background pancreas were manually segmented, and 488 radiomics features were extracted. Random forest classification based on radiomics features, age, and gender was evaluated with fourfold cross-validation. An academic radiologist blinded to the final pathologic diagnosis reviewed each case and provided the most likely diagnosis. RESULTS: 214 patients were included (64 intraductal papillary mucinous neoplasms, 33 mucinous cystic neoplasms, 60 serous cystadenomas, 24 solid pseudopapillary neoplasms, and 33 cystic neuroendocrine tumors). The radiomics-based machine learning approach showed AUC of 0.940 in pancreatic cyst classification, compared with AUC of 0.895 for the radiologist. CONCLUSION: Radiomics-based machine learning achieved equivalent performance as an experienced academic radiologist in the classification of pancreatic cysts. The high diagnostic accuracy can potentially maximize the efficiency of healthcare utilization by maximizing detection of high-risk lesions.

CT Radiomics-Based Preoperative Survival Prediction in Patients With Pancreatic Ductal Adenocarcinoma.

OBJECTIVE. Pancreatic ductal adenocarcinoma (PDAC) is often a lethal malignancy with limited preoperative predictors of long-term survival. The purpose of this study was to evaluate the prognostic utility of preoperative CT radiomics features in predicting postoperative survival of patients with PDAC. MATERIALS AND METHODS. A total of 153 patients with surgically resected PDAC who underwent preoperative CT between 2011 and 2017 were retrospectively identified. Demographic, clinical, and survival information was collected from the medical records. Survival time after the surgical resection was used to stratify patients into a low-risk group (survival time > 3 years) and a high-risk group (survival time < 1 year). The 3D volume of the whole pancreatic tumor and background pancreas were manually segmented. A total of 478 radiomics features were extracted from tumors and 11 extra features were computed from pancreas boundaries. The 10 most relevant features were selected by feature reduction. Survival analysis was performed on the basis of clinical parameters both with and without the addition of the selected features. Survival status and time were estimated by a random survival forest algorithm. Concordance index (C-index) was used to evaluate performance of the survival prediction model. RESULTS. The mean age of patients with PDAC was 67 ± 11 (SD) years. The mean tumor size was 3.31 ± 2.55 cm. The 10 most relevant radiomics features showed 82.2% accuracy in the classification of high-risk versus low-risk groups. The C-index of survival prediction with clinical parameters alone was 0.6785. The addition of CT radiomics features improved the C-index to 0.7414. CONCLUSION. Addition of CT radiomics features to standard clinical factors improves survival prediction in patients with PDAC.

Hepatic angiosarcoma in an adult who had Wilms tumor treated in childhood: A case report.

Hepatic angiosarcoma is a rare, highly aggressive mesenchymal liver malignancy with poor prognosis that stems from the endothelial cells that line the walls of blood or lymphatic vessels. It is the third most common primary liver malignancy and is most prevalent among older males. It is difficult to diagnose due to various clinical presentations from asymptomatic to abdominal pain, pleural effusion, and liver failure. The diagnosis of liver angiosarcoma is suspected on imaging features and confirmed by histopathological assessment. Primary management is determined based on the stage of tumor from surgery to palliative care such as chemotherapy or tumor transarterial embolization. We report a 51-year-old female who presented with stage 4 liver angiosarcoma and a history of childhood Wilms tumor. We focus on tumor management using radiological modalities and pathological analysis and discuss secondary liver tumors in survivors of childhood Wilms tumor.

Intra-abdominal lymphangioma: A case report.

Lymphangioma is a rare, benign congenital malformation of the lymphatic system that usually affects the neck and head in children. Intra-abdominal lymphangioma accounts for less than 5 percent of all cases of lymphangioma. The clinical presentation of intra-abdominal lymphangioma can vary from asymptomatic to nausea, vomiting, and abdominal pain. The diagnosis of intra-abdominal lymphangioma is based on imaging modalities and histopathological examination. The definitive treatment is surgical resection. Here we describe the interesting and rare case of a 29-year-old woman with lymphangioma of the retroperitoneum extending to the root of the mesentery. We focus on the diagnosis and management of this rare tumor by the application of radiological modalities and pathological analysis.

Pancreatic cystic desmoid tumor following metastatic colon cancer surgery: A case report.

Desmoid tumors are rare, benign, and locally aggressive neoplasms that stem from connective tissue that have high rates of recurrence after surgery. Intra-abdominal desmoid-type fibromatosis can arise in 2 forms: sporadic or hereditary (associated with familial adenomatous polyposis and Gardner syndrome). The diagnosis of desmoid-type tumors is based on imaging modalities and histopathological examination. The primary treatment is resection surgery. We report a 64-year-old male with a distal pancreatic desmoid tumor. We focus on tumor management by the application of radiological modalities and pathological analysis.

Idiopathic recurrent acute pancreatitis in the context of pancreas divisum: A case report.

Recurrent acute pancreatitis is characterized by frequent episodes of inflammation in the pancreas. The most common causes are alcohol abuse and gallstones but approximately 30% of cases are defined as idiopathic because initial evaluation fails to detect the etiology. In these patients, extensive laboratory and imaging evaluations usually lead to the uncovering of an occult biliary duct stone or sphincter of Oddi dysfunction as the main reason for frequent pancreatitis. We report a 42-year-old female with a longstanding history of acute recurrent pancreatitis in the setting of pancreas divisum due to its rarity. We focus on appropriate investigations by application of specialized laboratory and radiological modalities because determining the etiology is a crucial step in patient management.

Ectopic Cushing syndrome caused by a pancreatic neuroendocrine tumor: A case report.

Cushing syndrome is a disorder that occurs when the body is exposed to a higher than normal level of the hormone cortisol. It is most commonly caused by exogenous glucocorticoids, and less commonly due to endogenous sources. Ectopic adrenocorticotropic hormone (ACTH) syndrome is one of the rare causes of endogenous Cushing syndrome engendered by oversecretion of ACTH from a tumor outside of the pituitary or adrenal glands. We present a case of a 74-year-old male with uncontrolled type 2 diabetes mellitus who was suspected of having Cushing syndrome on chest CT due to increasing mediastinal lipomatosis and enlarging bilateral adrenal glands. Cushing syndrome was confirmed based on clinical features and biochemical tests. Further investigation revealed an ACTH-producing functional neuroendocrine tumor of the pancreas causing ectopic Cushing syndrome.

Uterine artery pseudoaneurysm following cesarean section; a case report.

Uterine artery pseudoaneurysm is an uncommon cause of vaginal bleeding that can occur after cesarean, hysterectomy, myomectomy, uncomplicated vaginal delivery, as well as gynecologic surgery. A 29-year-old woman (G4P1122) who underwent stat cesarean section and intrauterine device placement was found to have a 1.6-cm left uterine artery pseudoaneurysm on follow-up ultrasound. The patient presented to interventional radiology for angiography and uterine artery embolization to minimize the risk of spontaneous hemorrhage. This is the rare reported case, to our knowledge, of a uterine artery pseudoaneurysm associated with a recent cesarean section. Computed tomography angiogram, with multiplanar and maximal intensity projection images, can optimally display the pseudoaneurysm and the feeding vessel, which can provide valuable information for image-guided catheter embolization.

Clinical and histopathological outcome of cervical and chest MRI involving non-MRI-conditional cardiac pacemakers: a study using sheep models.

AIM: To examine the clinical and histopathological consequences of MRI in sheep implanted with non-MRI-conditional cardiac pacemakers. MATERIALS AND METHODS: Under general anesthesia, active fixation leads of two dual-chamber, non-MRI-conditional cardiac pacemakers (St. Jude Medical and Medtronic) were implanted either at the right ventricular apex or at the right atrium of two male sheep and connected to the V and A channels of the pacemakers, respectively. The generators were placed in cervical subcutaneous pockets. On day 5, both sheep underwent 1.5 T cervical and chest MRI with continuous electrocardiogram monitoring. Obtained sequences were T1-weighted (T1W), T2-weighted (T2W), T2-gradient echo and diffusion weighted (DW). The employed modes were OVO, VOO and VVI for one sheep and OAO, AOO and AAI for the other (unipolar and bipolar configuration of pacing and sensing for both). Battery impedance, pacing lead impedance, intrinsic amplitude and capture thresholds were checked at baseline and after each sequence, as well as 48 h after imaging. Histopathological examination of the cardiac tissue around the lead tip was performed 4 weeks post-imaging. RESULTS: No significant changes in device position or configuration were observed during or after MRI. Clinical outcome was uneventful in both sheep. Minor inflammatory and necrotic changes were reported after histopathological examination of the cardiac tissue around the lead tip. CONCLUSION: 1.5 T MRI of two implanted non-MRI-conditional pacemakers was found safe in terms of device configuration and stability, clinical outcome and cardiac tissue histopathological findings.

Accuracy of ultrasonography in detecting gastrointestinal tract obstructions in children; a prospective study in a tertiary referral hospital.

AIMS: To examine the diagnostic accuracy of plain radiography, abdominal ultrasonography (US), and their combination in pediatric patients with suspected gastrointestinal (GI) tract obstruction. MATERIAL AND METHODS: A cohort of 48 patients (age, 0-14 years, 27 boys) with clinical manifestations of GI tract obstruction underwent plain radiography and abdominal US examination. The final diagnoses were based on intraoperative findings, rectal biopsies (in Hirschsprung's disease), or adequate follow-ups. RESULTS: The GI tract obstruction was diagnosed in 40 patients. The sensitivity, specificity, positive predictive value and negative predictive value of plain radiography in diagnosing GI tract obstruction were 87.5%, 75.0%, 94.6%, and 54.6%, respectively. The corresponding values were 95%, 100%, 100%, and 80%, respectively when US was used alone; and 97.5%, 100%, 100% and 88.9%, respectively when radiography and US were used together. Except for two patients (one with Hirschsprung's disease and the other with massive peritonitis), US detected the underlying causes of obstruction correctly in all patients. CONCLUSIONS: US is a highly sensitive and specific modality in diagnosing pediatric GI tract obstructions, as well as their causes. The combination of plain radiography and US further increase the diagnostic sensitivity and negative predictive value.

Invasive Intraductal Papillary Mucinous Neoplasms: CT Features of Colloid Carcinoma Versus Tubular Adenocarcinoma of the Pancreas.

OBJECTIVE. The purpose of this study is to compare the CT features of colloid carcinoma and tubular adenocarcinoma of the pancreas arising in association with intraductal papillary mucinous neoplasms (IPMNs). MATERIALS AND METHODS. The preoperative CT images of 85 patients with histopathologically proven IPMNs and associated invasive adenocarcinoma located next to each other were retrospectively reviewed. Twenty-nine patients (34.1%; 19 men and 10 women; mean [± SD] age, 68.0 ± 9.5 years) had invasive colloid carcinoma, and 56 patients (65.9%; 31 men and 25 women; mean age, 70.8 ± 10.6 years) had invasive tubular adenocarcinoma. We compared the following CT features between the two groups: IPMN type, main pancreatic duct (MPD) and common bile duct diameters, diameter and characteristics of the largest cystic lesion for branch duct and mixed-type IPMNs, presence of an extracystic or extraductal solid mass next to the cystic lesion or MPD, morphologic features of the upstream MPD in relation to the cystic lesion or solid mass, and presence of a fistula to the adjacent organs. RESULTS. An MPD size of 9.5 mm or greater, a largest cystic lesion diameter of 28 mm or greater, location in the head or neck, septation, calcification, presence of a mural nodule(s) within a cystic lesion or MPD, and presence of a fistula were all more commonly associated with colloid carcinoma. In contrast, presence of an extracystic or extraductal solid mass and an abrupt change in the caliber of the dilated MPD were associated with tubular adenocarcinoma. The best CT feature for differentiating between the two groups was the morphologic features of the upstream MPD in relation to the cystic lesion or solid mass (sensitivity, 81.3%; specificity, 92.3%). CONCLUSION. Preoperative CT is helpful in differentiating two types of invasive carcinoma arising in association with IPMNs. These findings are clinically important because prognosis is better for colloid carcinoma than for tubular adenocarcinoma.

Diagnostic performance of 64-MDCT in detecting ERCP-proven periampullary duodenal diverticula.

AIM: To determine the diagnostic performance of 64-slice multidetector computed tomography (64-MDCT) in detecting periampullary duodenal diverticula. MATERIALS AND METHODS: Medical profiles of 120 endoscopic retrograde cholangiopancreatography (ERCP)-proven patients with (n = 100) and without (n = 20) periampullary duodenal diverticula who had undergone 64-MDCT were retrospectively reviewed. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 64-MDCT in detecting periampullary duodenal diverticula were calculated. Potential factors that might influence the diagnostic performance of 64-MDCT in such patients were also examined. RESULTS: Patients were 60 males and 60 females with the mean age of 68.8 ± 12.7 (27-93) years. Indications of ERCP were common bile duct stricture (n = 62) or stone (n = 41), biliary cholestasis (n = 16) and acute cholangitis (n = 1). The sensitivity, specificity, PPV, and NPV of 64-MDCT in detecting periampullary duodenal diverticula were 76%, 100%, 100%, and 45.5%, respectively. The size of diverticula was the only predictor of 64-MDCT performance, with better results observed in larger (> 20 mm) diverticula. CONCLUSION: 64-MDCT is a highly specific imaging modality in detecting periampullary duodenal diverticula. The diagnostic performance of 64-MDCT increases for larger diverticula.

CT imaging, classification, and complications of acute pancreatitis.

Acute pancreatitis is an increasingly common condition and can result in significant morbidity and mortality. Contrast enhanced computed tomography (CECT) is the primary initial imaging modality in the characterization of acute pancreatitis. In this article, we provide sample CECT technical acquisition parameters for pancreatic imaging. We also review the classification systems for acute pancreatitis and give examples of common and uncommon complications of acute pancreatitis.

Baseline 3D-ADC outperforms 2D-ADC in predicting response to treatment in patients with colorectal liver metastases.

OBJECTIVES: To examine the value of baseline 3D-ADC and to predict short-term response to treatment in patients with hepatic colorectal metastases (CLMs). METHODS: Liver MR images of 546 patients with CLMs (2008-2015) were reviewed retrospectively and 68 patients fulfilled inclusion criteria. Patients had received systemic chemotherapy (n = 17), hepatic trans-arterial chemoembolization or TACE (n = 34), and 90Y radioembolization (n = 17). Baseline (pre-treatment) 3D-ADC (volumetric) of metastatic lesions was calculated employing prototype software. RECIST 1.1 was used to assess short-term response to treatment. Prediction of response to treatment by baseline 3D-ADC and 2D-ADC (ROI-based) was also compared in all patients. RESULTS: Partial response to treatment (minimum 30% decrease in tumor largest transverse diameter) was seen in 35.3% of patients; 41.2% with systemic chemotherapy, 32.4% with TACE, and 35.3% with 90Y radioembolization (p = 0.82). Median baseline 3D-ADC was significantly lower in responding than in nonresponding lesions. Area under the curve (AUC) of 3D-ADC was 0.90 in 90Y radioembolization patients, 0.88 in TACE patients, and 0.77 in systemic chemotherapy patients (p < 0.01). Optimal prediction was observed with the 10th percentile of ADC (1006 × 10-6 mm2/s), yielding sensitivity and specificity of 77.4% and 91.3%, respectively. 3D-ADC outperformed 2D-ADC in predicting response to treatment (AUC; 0.86 vs. 0.71; p < 0.001). CONCLUSION: Baseline 3D-ADC is a highly specific biomarker in predicting partial short-term response to treatment in hepatic CLMs. KEY POINTS: • Baseline 3D-ADC is a highly specific biomarker in predicting response to different treatments in hepatic CLMs. • The prediction level of baseline ADC is better for90Y radioembolization than for systemic chemotherapy/TACE in hepatic CLMs. • 3D-ADC outperforms 2D-ADC in predicting short-term response to treatment in hepatic CLMs.

Emerging imaging techniques for acute pancreatitis.

Acute pancreatitis (AP) is caused by acute inflammation of the pancreas and adjacent tissue and is a common source of abdominal pain. The current CT and MRI evaluation of AP is mostly based on morphologic features. Recent advances in image acquisition and analysis offer the opportunity to go beyond morphologic features. Advanced MR techniques such as diffusion-weighted imaging, as well as T1 and T2 mapping, can potentially quantify signal changes reflective of underlying tissue abnormalities. Advanced analytic techniques such as radiomics and artificial neural networks (ANNs) offer the promise of uncovering imaging biomarkers that can provide additional classification and prognostic information. The purpose of this article is to review recent advances in imaging acquisition and analytic techniques in the evaluation of AP.

Urinary bladder fistulae and the role of CT cystography: a pictorial review.

Various conditions such as inflammation, malignancy, surgical manipulations, and radiation therapy can lead to the development of urinary bladder fistulae. Although many of these fistulae are suspected clinically, imaging plays a major role in confirmation and planning of subsequent treatment. Computed tomography (CT) cystoscopy provides useful details regarding the fistulous track and the anatomy of the region. This article aims to provide a succinct review of bladder fistulae and the role of CT cystography in their management.

Imaging of urinary bladder injury: the role of CT cystography.

Although conventional radiographic cystography has been traditionally considered the reference standard in detecting bladder injuries, computed tomography (CT) cystography has become the initial imaging method of choice in the acute setting. CT cystography has been shown to provide comparable accuracy as conventional cystography, and can be easily performed in conjunction with trauma CT surveys in patients with suspected bladder injuries. Despite increasing enthusiasm toward CT cystography in dealing with patients with suspected bladder injuries, there is little information in this regard in the literature. This article aims to discuss the role of CT cystography in the evaluation of bladder injuries.