Preoperative Prediction of Lymph Node Metastases in Nonfunctional Pancreatic Neuroendocrine Tumors Using a Combined CT Radiomics-Clinical Model.

BACKGROUND: PanNETs are a rare group of pancreatic tumors that display heterogeneous histopathological and clinical behavior. Nodal disease has been established as one of the strongest predictors of patient outcomes in PanNETs. Lack of accurate preoperative assessment of nodal disease is a major limitation in the management of these patients, in particular those with small (< 2 cm) low-grade tumors. The aim of the study was to evaluate the ability of radiomic features (RF) to preoperatively predict the presence of nodal disease in pancreatic neuroendocrine tumors (PanNETs). PATIENTS AND METHODS: An institutional database was used to identify patients with nonfunctional PanNETs undergoing resection. Pancreas protocol computed tomography was obtained, manually segmented, and RF were extracted. These were analyzed using the minimum redundancy maximum relevance analysis for hierarchical feature selection. Youden index was used to identify the optimal cutoff for predicting nodal disease. A random forest prediction model was trained using RF and clinicopathological characteristics and validated internally. RESULTS: Of the 320 patients included in the study, 92 (28.8%) had nodal disease based on histopathological assessment of the surgical specimen. A radiomic signature based on ten selected RF was developed. Clinicopathological characteristics predictive of nodal disease included tumor grade and size. Upon internal validation the combined radiomics and clinical feature model demonstrated adequate performance (AUC 0.80) in identifying nodal disease. The model accurately identified nodal disease in 85% of patients with small tumors (< 2 cm). CONCLUSIONS: Non-invasive preoperative assessment of nodal disease using RF and clinicopathological characteristics is feasible.

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.

Radiologists' Expectations of Artificial Intelligence in Pancreatic Cancer Imaging: How Good Is Good Enough?

BACKGROUND: Existing (artificial intelligence [AI]) tools in radiology are modeled without necessarily considering the expectations and experience of the end user-the radiologist. The literature is scarce on the tangible parameters that AI capabilities need to meet for radiologists to consider them useful tools. OBJECTIVE: The purpose of this study is to explore radiologists' attitudes toward AI tools in pancreatic cancer imaging and to quantitatively assess their expectations of these tools. METHODS: A link to the survey was posted on the www.ctisus.com website, advertised in the www.ctisus.com email newsletter, and publicized on LinkedIn, Facebook, and Twitter accounts. This survey asked participants about their demographics, practice, and current attitudes toward AI. They were also asked about their expectations of what constitutes a clinically useful AI tool. The survey consisted of 17 questions, which included 9 multiple choice questions, 2 Likert scale questions, 4 binary (yes/no) questions, 1 rank order question, and 1 free text question. RESULTS: A total of 161 respondents completed the survey, yielding a response rate of 46.3% of the total 348 clicks on the survey link. The minimum acceptable sensitivity of an AI program for the detection of pancreatic cancer chosen by most respondents was either 90% or 95% at a specificity of 95%. The minimum size of pancreatic cancer that most respondents would find an AI useful at detecting was 5 mm. Respondents preferred AI tools that demonstrated greater sensitivity over those with greater specificity. Over half of respondents anticipated incorporating AI tools into their clinical practice within the next 5 years. CONCLUSION: Radiologists are open to the idea of integrating AI-based tools and have high expectations regarding the performance of these tools. Consideration of radiologists' input is important to contextualize expectations and optimize clinical adoption of existing and future AI tools.

The Impact of Clinical and Pathological Features on Intraductal Papillary Mucinous Neoplasm Recurrence After Surgical Resection: Long-Term Follow-Up Analysis.

OBJECTIVE: This study aimed to identify risk factors for recurrence after pancreatic resection for intraductal papillary mucinous neoplasm (IPMN). SUMMARY BACKGROUND DATA: Long-term follow-up data on recurrence after surgical resection for IPMN are currently lacking. Previous studies have presented mixed results on the role of margin status in risk of recurrence after surgical resection. METHODS: A total of 126 patients that underwent resection for noninvasive IPMN were followed for a median of 9.5 years. Dedicated pathological and radiological reviews were performed to correlate clinical and pathological features (including detailed pathological features of the parenchymal margin) with recurrence after surgical resection. In addition, in a subset of 32 patients with positive margins, we determined the relationship between the margin and original IPMN using driver gene mutations identified by next-generation sequencing. RESULTS: Family history of pancreatic cancer and high-grade IPMN was identified as risk factors for recurrence in both uni- and multivariate analysis (adjusted hazard ratio 3.05 and 1.88, respectively). Although positive margin was not significantly associated with recurrence in our cohort, the size and grade of the dysplastic focus at the margin were significantly correlated with recurrence in margin-positive patients. Genetic analyses showed that the neoplastic epithelium at the margin was independent from the original IPMN in at least 9 of 32 cases (28%). The majority of recurrences (74%) occurred after 3 years, and a significant minority (32%) occurred after 5 years. CONCLUSION: Sustained postoperative surveillance for all patients is indicated, particularly those with risk factors such has family history and high-grade dysplasia.

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.

Current Status of Radiomics and Deep Learning in Liver Imaging.

ABSTRACT: Artificial intelligence is poised to revolutionize medical image. It takes advantage of the high-dimensional quantitative features present in medical images that may not be fully appreciated by humans. Artificial intelligence has the potential to facilitate automatic organ segmentation, disease detection and characterization, and prediction of disease recurrence. This article reviews the current status of artificial intelligence in liver imaging and reviews the opportunities and challenges in clinical implementation.

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.

Utility of CT Radiomics Features in Differentiation of Pancreatic Ductal Adenocarcinoma From Normal Pancreatic Tissue.

OBJECTIVE. The objective of our study was to determine the utility of radiomics features in differentiating CT cases of pancreatic ductal adenocarcinoma (PDAC) from normal pancreas. MATERIALS AND METHODS. In this retrospective case-control study, 190 patients with PDAC (97 men, 93 women; mean age ± SD, 66 ± 9 years) from 2012 to 2017 and 190 healthy potential renal donors (96 men, 94 women; mean age ± SD, 52 ± 8 years) without known pancreatic disease from 2005 to 2009 were identified from radiology and pathology databases. The 3D volume of the pancreas was manually segmented from the preoperative CT scans by four trained researchers and verified by three abdominal radiologists. Four hundred seventy-eight radiomics features were extracted to express the phenotype of the pancreas. Forty features were selected for analysis because of redundancy of computed features. The dataset was divided into 255 training cases (125 normal control cases and 130 PDAC cases) and 125 validation cases (65 normal control cases and 60 PDAC cases). A random forest classifier was used for binary classification of PDAC versus normal pancreas of control cases. Accuracy, sensitivity, and specificity were calculated. RESULTS. Mean tumor size was 4.1 ± 1.7 (SD) cm. The overall accuracy of the random forest binary classification was 99.2% (124/125), and AUC was 99.9%. All PDAC cases (60/60) were correctly classified. One case from a renal donor was misclassified as PDAC (1/65). The sensitivity was 100%, and specificity was 98.5%. CONCLUSION. Radiomics features extracted from whole pancreas can be used to differentiate between CT cases from patients with PDAC and healthy control subjects with normal pancreas.

TIRADS Interobserver Variability Among Indeterminate Thyroid Nodules: A Single-Institution Study.

OBJECTIVES: A high proportion of cytologically indeterminate, Afirma Gene Expression Classifier "suspicious" thyroid nodules are benign. The Thyroid Imaging Reporting and Data System (TIRADS), was proposed by the American College of Radiology in 2017 to help classify thyroid nodules based on ultrasound characteristics in a standardized fashion to guide management. We aim to determine the interobserver variability of TIRADS classification among cytologically indeterminate and Afirma suspicious nodules. METHODS: We retrospectively queried cytopathology archives for thyroid fine-needle aspiration specimens obtained between February 2012 and September 2016 with associated (1) indeterminate diagnosis, (2) ultrasound imaging at our institution, (3) Afirma suspicious result, and (4) surgery at our institution. We compared the TIRADS variability of the 3 blinded radiologists using intraclass correlation coefficients. RESULTS: Our cohort consisted of 127 nodules. Intraclass correlation coefficients can be interpreted as follows: less than 0.4, poor; 0.4 to 0.59, fair; 0.6 to 0.74, good; 0.75 to 1.00, excellent. The intraclass correlation coefficients of the raw TIRADS score and category variability was 0.561 (95% confidence interval [CI]: 0.464-0.651) or fair and 0.547 (95% CI, 0.449-0.640) or fair, respectively. When analyzing composition, echogenicity, shape, margin, and echogenic foci, the ICCs were 0.552 (95% CI, 0.454-0.643), fair; 0.533 (95% CI, 0.432-0.627), fair; 0.359 (95% CI, 0.248-0.469), poor; 0.192 (95% CI, 0.084-0.308), poor; and 0.549 (95% CI, 0.451- 0.641), fair, respectively. CONCLUSIONS: Our results show that among the subset of cytologically indeterminate and Afirma suspicious nodules, TIRADS interobserver variability was fair. Shape and margin criteria were the biggest sources of disagreement. Large prospective studies are needed to evaluate the interobserver variability of TIRADS in this subset of thyroid nodules.

CT Detection of Symptomatic and Asymptomatic Meckel Diverticulum.

OBJECTIVE: The objective of our study was to determine how often symptomatic Meckel diverticulum and asymptomatic Meckel diverticulum are detected on CT in patients with known Meckel diverticulum and to evaluate factors that influence detection. MATERIALS AND METHODS: A total of 85 CT examinations of 40 patients (eight pediatric patients and 32 adult patients; 29 male patients and 11 female patients; average age, 46.2 ± 23.7 [SD] years) with a pathologic diagnosis of Meckel diverticulum were retrospectively evaluated. These patients included 26 adult patients with incidentally found asymptomatic Meckel diverticulum and 14 patients (eight pediatric and six adult patients) with symptomatic Meckel diverticulum. The CT technical factors and patients' morphologic factors were compared with the detection of Meckel diverticulum using mixed-effect logistic regression models. RESULTS: Meckel diverticulum was detected on at least one CT examination in eight of 14 (57.1%) symptomatic patients (two of four patients with bleeding, two of six patients with small-bowel obstruction, two of two patients with acute diverticulitis, one of one patient with incisional hernia, and one of one patient with inverted Meckel diverticulum) and in 13 of 23 (56.5%) total CT examinations. Asymptomatic Meckel diverticulum was detected on at least one CT examination in 11 of 26 (42.3%) patients and in 16 of 62 (25.8%) total CT examinations. The amount of peritoneal fat was related to the detection of Meckel diverticula (p = 0.02). Although not statistically significant, the subjective quality of axial CT (p = 0.05) tended to be related to detection, whereas the use of IV (p = 0.59) or positive oral (p = 0.41) contrast material was unrelated to detection. In the original CT reports, none of the asymptomatic cases of Meckel diverticulum was prospectively detected, whereas Meckel diverticulum was detected or mentioned as a possibility in 64% of the symptomatic cases. CONCLUSION: In patients with known Meckel diverticulum, CT can detect Meckel diverticulum in up to 47.5% of all patients. Meckel diverticulum is more commonly detected in symptomatic patients than in asymptomatic patients, and detection is related to the amount of peritoneal fat.

Preliminary data using computed tomography texture analysis for the classification of hypervascular liver lesions: generation of a predictive model on the basis of quantitative spatial frequency measurements--a work in progress.

OBJECTIVE: Computed tomography texture analysis (CTTA) is a method of quantifying lesion heterogeneity based on distribution of pixel intensities within a region of interest. This study investigates the ability of CTTA to distinguish different hypervascular liver lesions and compares CTTA parameters by creating a proof-of-concept model to distinguish between different lesions. METHODS: Following institutional review board approval, CTTA software (TexRAD Ltd) was used to retrospectively analyze 17 cases of focal nodular hyperplasia, 19 hepatic adenomas, 25 hepatocellular carcinomas, and 19 cases of normal liver parenchyma using arterial phase scans. Two radiologists read the same image series used by the CTTA software and reported their best guess diagnosis. Computed tomography texture analysis parameters were computed from regions of interest using spatial band-pass filters to quantify heterogeneity. Random-forest method was used to construct a predictive model from these parameters, and a separate regression model was created using a subset of parameters. RESULTS: The random-forest model successfully distinguished the 3 lesion types and normal liver with predicted classification performance accuracy for 91.2% for adenoma, 94.4% for focal nodular hyperplasia, and 98.6% for hepatocellular carcinoma. This error prediction was generated using a subset of data points not used in generation of the model, but not on discrete prospective cases. In contrast, the 2 human readers using the same image series data analyzed by the CTTA software had lower accuracies, of 72.2% and 65.6%, respectively. The explicit regression model with a subset of image parameters had intermediate overall accuracy of 84.9%. CONCLUSIONS: Computed tomography texture analysis may prove valuable in lesion characterization. Differentiation between common hypervascular lesion types could be aided by the judicious incorporation of texture parameters into clinical analysis.

Diagnostic performance of multidetector computed tomography in distinguishing unilateral from bilateral abnormalities in primary hyperaldosteronism: comparison of multidetector computed tomography with adrenal vein sampling.

OBJECTIVE: The management of patients with primary hyperaldosteronism (PH) varies depending on whether the unregulated aldosterone secretion localizes to a single unilateral adrenal gland, traditionally determined using adrenal vein sampling (AVS). This study seeks to determine if the performance of multidetector computed tomography (MDCT) examinations performed using the latest scanner technology can reasonably match the results of AVS, and potentially avoid AVS in some patients. MATERIALS AND METHODS: Computed tomographic scans in 56 patients with PH were independently reviewed by 2 radiologists for the presence of adrenal nodules and qualitative adrenal thickening. Results were correlated with AVS results. RESULTS: Of 35 patients with MDCT evidence of unilateral nodules, the imaging findings correctly predicted AVS localization in only 23 (65.7%) cases. When stratified by size, MDCT was accurate in only 71.4% of cases for nodules measuring 10 mm or less, and only 55.0% of cases for nodules measuring 11 to 20 mm. Of the 12 cases where MDCT did not correctly localize, AVS localized to the contralateral adrenal gland in 4 cases, whereas AVS suggested no lateralization in 8 cases. In patients with normal bilateral adrenal glands on MDCT, 2/7 (28.6%) of cases demonstrated unilateral localization on AVS, and in patients with bilateral adrenal nodules, only 3/14 (21.4%) did not demonstrate lateralization on AVS. CONCLUSIONS: Multidetector computed tomography, even when performed with the latest generation of MDCT scanners, does not offer sufficient diagnostic accuracy to replace AVS in patients with PH.

Duodenal neuroendocrine tumors: retrospective evaluation of CT imaging features and pattern of metastatic disease on dual-phase MDCT with pathologic correlation.

PURPOSE: The purpose of the study is to evaluate the CT appearance and pattern of metastatic disease of patients with surgically resected well-differentiated duodenal neuroendocrine tumors who underwent pre-operative dual-phase CT. METHODS: Clinical and pathologic records and CT images of 28 patients (average age 58.0 years) following Whipple procedure were retrospectively reviewed. The size, morphology (polypoid, intraluminal mass or wall thickening, intramural mass), location, CT attenuation in the arterial and venous phases, and the presence of lymph node or liver metastases were recorded. RESULTS: On CT, 19 patients (67.8%) had neuroendocrine tumors manifested as polypoid or intraluminal masses (38 lesions, multiple tumors in 3 patients), 4 patients (14.3%) had tumors manifested as wall thickening or intramural masses, and in 5 patients (17.9%), the primary tumor was not visualized. Lesions not seen at CT were less than 0.8 cm on pathologic diagnosis. The mean size of polypoid tumors on CT was 1.2 cm (range 0.3-3.8 cm); 24 tumors were 1.0 cm or smaller, and 14 tumors were larger than 1.0 cm. Most lesions were hypervascular in the arterial phase (19/23 patients) with an increase in tumor enhancement in the venous phase in 14 patients (60.9%), decrease in enhancement in 7 patients (30.4%), and no change in enhancement in 2 patients (8.7%). Thirteen patients (46.4%) had metastatic disease from carcinoid tumor, most commonly regional enhancing lymphadenopathy. CONCLUSION: Duodenal carcinoid tumors commonly appear as an enhancing mass in either the arterial or venous phases. If a primary tumor is not seen in the duodenum, adjacent enhancing lymphadenopathy can be a clue to the presence of a duodenal carcinoid tumor.

CT and MRI assessment of symptomatic organized pancreatic fluid collections and pancreatic duct disruption: an interreader variability study using the revised Atlanta classification 2012.

PURPOSE: Compare CT and MRI for fluid/debris component estimate and pancreatic duct (PD) communication with organized pancreatic fluid collections in acute pancreatitis. Evaluate fat density globules on CT as marker for debris. METHODS: 29 Patients with 46 collections with CECT and MRI performed ≥4 weeks of symptom onset assessed for necrotizing pancreatitis, estimated percentage of fluid volume and PD involvement by two radiologists on separate occasions. T2WI used as standard for estimated percentage of fluid volume. Presence of fat globules and fluid attenuation on CT was recorded. Spearman rank correlation and kappa statistics were used to assess the correlation between imaging techniques and interreader agreement, respectively. RESULTS: Necrotizing pancreatitis seen on CT in 27 (93%, κ 0.119) vs. 20 (69%, κ 0.748) patients on MRI. CT identified 42 WON and 4 pseudocysts vs. 34 WON, and 12 pseudocysts on MRI. Higher interreader agreement for percentage fluid volume on MRI (κ = 0.55) vs. CT (κ = 0.196). Accuracy of CT in evaluation of percentage fluid volume was 65% using T2WI MRI used as standard. Fat globules identified on CT in 13(65%) out of 20 collections containing <75% fluid vs. 4(15%) out of 26 collections containing >75% fluid (p = 0.0001). PD involvement confidently excluded on CT in 68% collections vs. 93% on MRI. CONCLUSION: MRI demonstrates higher reproducibility for fluid to debris component estimation. Fat globules on CT were frequently seen in organized pancreatic fluid collections with large amount of debris. PD disruption more confidently excluded on MRI. This information may be helpful for pre-procedure planning.

Differentiating autoimmune pancreatitis from pancreatic adenocarcinoma using dual-phase computed tomography.

OBJECTIVE: This article aimed to study features on dual-phase computed tomography (CT) that help differentiate autoimmune pancreatitis (AIP) from pancreatic adenocarcinoma (PA). METHODS: The CTs of 32 patients with AIP were matched with equal number of PA and were independently evaluated by 3 radiologists who assigned a diagnosis of AIP, PA, or unsure. Interobserver agreement between radiologists was evaluated using κ statistics. RESULTS: The mean accuracies for diagnosing AIP and PA were 68% and 83%, respectively. There was moderate agreement between radiologists (κ, 0.58; P < 0.0001). The most common findings for AIP were common bile duct (CBD) stricture (63%), bile duct wall hyperenhancement (47%), and diffuse parenchymal enlargement (41%). The most common findings for PA were focal mass (78%; κ, 0.58; P < 0.0001) and pancreatic ductal dilatation (69%; κ, 0.7; P < 0.0001). Findings helpful for diagnosing AIP were diffuse enlargement, parenchymal atrophy as well as absence of pancreatic duct dilatation and focal mass. Findings helpful for diagnosing PA were focal mass and pancreatic ductal dilatation. Misdiagnosis of PA in patients with AIP was due to focal mass, pancreatic duct dilatation, and pancreatic atrophy, whereas misdiagnosis of AIP in patients with PA was due to absence of atrophy, presence of diffuse enlargement, and peripancreatic halo. CONCLUSIONS: Diffuse enlargement, hypoenhancement, and characteristic peripancreatic halo are strong indicators for a diagnosis of AIP. Radiologists demonstrated moderate agreement in distinguishing AIP from PA on the basis of CT imaging.

Intrapancreatic accessory spleen: possibilities of computed tomography in differentiation from nonfunctioning pancreatic neuroendocrine tumor.

OBJECTIVE: The aim of this study was to evaluate the ability of computed tomography (CT) in differentiating between intrapancreatic accessory spleen (IPAS) from pancreatic neuroendocrine tumor (PanNET). METHODS: Eight IPASs and 12 PanNETs in the pancreatic tail were retrospectively evaluated by 2 radiologists. Readers assigned a diagnosis to each examination and evaluated for the presence or absence of 9 CT findings that may aid in the diagnosis. RESULTS: Reader 1 had a sensitivity of 0.83 and a specificity of 1; reader 2 had a sensitivity of 0.78 and a specificity of 0.86. Three of the 9 CT findings were found to be statistically significant in IPASs: the lesion present along the pancreatic dorsal surface, the lesion demonstrating the same enhancement as the spleen on venous phase, and heterogeneous enhancement during arterial phase. CONCLUSIONS: CT can be used to differentiate between IPAS and PanNET with good specificity and sensitivity. The IPAS mirrors the spleen's enhancement and is usually located along the dorsal surface of the pancreas.

Early detection of pancreatic cancer in the era of precision medicine.

Pancreatic ductal adenocarcinoma (PDAC) is the third leading cause of cancer-related mortality and it is often diagnosed at advanced stages due to non-specific clinical presentation. Disease detection at localized disease stage followed by surgical resection remains the only potentially curative treatment. In this era of precision medicine, a multifaceted approach to early detection of PDAC includes targeted screening in high-risk populations, serum biomarkers and "liquid biopsies", and artificial intelligence augmented tumor detection from radiologic examinations. In this review, we will review these emerging techniques in the early detection of PDAC.

Hidden in plain sight: commonly missed early signs of pancreatic cancer on CT.

Pancreatic ductal adenocarcinoma (PDAC) has poor prognosis mostly due to the advanced stage at which disease is diagnosed. Early detection of disease at a resectable stage is, therefore, critical for improving outcomes of patients. Prior studies have demonstrated that pancreatic abnormalities may be detected on CT in up to 38% of CT studies 5 years before clinical diagnosis of PDAC. In this review, we highlight commonly missed signs of early PDAC on CT. Broadly, these commonly missed signs consist of small isoattenuating PDAC without contour deformity, isolated pancreatic duct dilatation and cutoff, focal pancreatic enhancement and focal parenchymal atrophy, pancreatitis with underlying PDAC, and vascular encasement. Through providing commentary on demonstrative examples of these signs, we demonstrate how to reduce the risk of missing or misinterpreting radiological features of early PDAC.

Accurate non-invasive grading of nonfunctional pancreatic neuroendocrine tumors with a CT derived radiomics signature.

PURPOSE: The purpose of this study was to develop a radiomics-signature using computed tomography (CT) data for the preoperative prediction of grade of nonfunctional pancreatic neuroendocrine tumors (NF-PNETs). MATERIALS AND METHODS: A retrospective study was performed on patients undergoing resection for NF-PNETs between 2010 and 2019. A total of 2436 radiomic features were extracted from arterial and venous phases of pancreas-protocol CT examinations. Radiomic features that were associated with final pathologic grade observed in the surgical specimens were subjected to joint mutual information maximization for hierarchical feature selection and the development of the radiomic-signature. Youden-index was used to identify optimal cutoff for determining tumor grade. A random forest prediction model was trained and validated internally. The performance of this tool in predicting tumor grade was compared to that of EUS-FNA sampling that was used as the standard of reference. RESULTS: A total of 270 patients were included and a fusion radiomic-signature based on 10 selected features was developed using the development cohort (n = 201). There were 149 men and 121 women with a mean age of 59.4 ± 12.3 (standard deviation) years (range: 23.3-85.0 years). Upon internal validation in a new set of 69 patients, a strong discrimination was observed with an area under the curve (AUC) of 0.80 (95% confidence interval [CI]: 0.71-0.90) with corresponding sensitivity and specificity of 87.5% (95% CI: 79.7-95.3) and 73.3% (95% CI: 62.9-83.8) respectively. Of the study population, 143 patients (52.9%) underwent EUS-FNA. Biopsies were non-diagnostic in 26 patients (18.2%) and could not be graded due to insufficient sample in 42 patients (29.4%). In the cohort of 75 patients (52.4%) in whom biopsies were graded the radiomic-signature demonstrated not different AUC as compared to EUS-FNA (AUC: 0.69 vs. 0.67; P = 0.723), however greater sensitivity (i.e., ability to accurately identify G2/3 lesion was observed (80.8% vs. 42.3%; P < 0.001). CONCLUSION: Non-invasive assessment of tumor grade in patients with PNETs using the proposed radiomic-signature demonstrated high accuracy. Prospective validation and optimization could overcome the commonly experienced diagnostic uncertainty in the assessment of tumor grade in patients with PNETs and could facilitate clinical decision-making.

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.