Search

Purpose: Our study predictively maps epithelium density in magnetic resonance imaging (MRI) space while varying the ground truth labels provided by five pathologists to quantify the downstream effects of interobserver variability. Approach: Clinical imaging and postsurgical tissue from 48 recruited prospective patients were used in our study. Tissue was sliced to match the MRI orientation and whole-mount slides were stained and digitized. Data from 28 patients (n = 33 slides) were sent to five pathologists to be annotated. Slides from the remaining 20 patients (n = 123 slides) were annotated by one of the five pathologists. Interpathologist variability was measured using Krippendorff’s alpha. Pathologist-specific radiopathomic mapping models were trained using a partial least-squares regression using MRI values to predict epithelium density, a known marker for disease severity. An analysis of variance characterized intermodel means difference in epithelium density. A consensus model was created and evaluated using a receiver operator characteristic classifying high grade versus low grade and benign, and was statistically compared to apparent diffusion coefficient (ADC). Results: Interobserver variability ranged from low to acceptable agreement (0.31 to 0.69). There was a statistically significant difference in mean predicted epithelium density values (p

Purpose: This study aims to combine multiparametric magnetic resonance imaging (MRI) and digitized pathology with machine learning to generate predictive maps of histologic features for prostate cancer localization. Methods and Materials: Thirty-nine patients underwent MRI prior to prostatectomy. After surgery, tissue was sliced according to MRI orientation using patient-specific 3-dimensionally printed slicing jigs. Whole-mount sections were annotated by our pathologist and digitally contoured to differentiate the lumen and epithelium. Slides were co-registered to the T2-weighted MRI scan. A learning curve was generated to determine the number of patients required for a stable machine-learning model. Patients were randomly stratified into 2 training sets and 1 test set. Two partial least-squares regression models were trained, each capable of predicting lumen and epithelium density. Predicted density values were calculated for each patient in the test dataset, mapped into the MRI space, and compared between regions confirmed as high-grade prostate cancer. Results: The learning-curve analysis showed that a stable fit was achieved with data from 10 patients. Maps indicated that regions of increased epithelium and decreased lumen density, generated from each independent model, corresponded with pathologist-annotated regions of high-grade cancer. Conclusions: We present a radio-pathomic approach to mapping prostate cancer. We find that the maps are useful for highlighting high-grade tumors. This technique may be relevant for dose-painting strategies in prostate radiation therapy. © 2018 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)., Summary This study aims to combine whole-mount prostate pathology with multiparametric magnetic resonance imaging from 39 patients to generate predictive maps of epithelium and lumen density in magnetic resonance imaging space. We show that the new image contrasts generated stratify high-grade tumors from low-grade tumors and healthy tissue. Future studies will explore targeted radiation therapy and clinical disease staging using the radio-pathomic mapping technique.

Selective screening for pancreatic cancer (PC) has been proposed. We describe the establishment of a comprehensive multidisciplinary screening program using 3.0 T MRI. Criteria for screening included the presence of PC in: ≥ 2 first degree relatives (FDR), 1 FDR and 1 s degree relative (SDR), ≥ 3 any degree relatives (ADR), or any known hereditary cancer syndrome with increased PC risk. Imaging with 3.0 T MRI was performed routinely and endoscopic ultrasound was used selectively. Screening was completed in 75 patients (pts). Hereditary cancer syndromes were present in 42 (56%) of the 75 pts: BRCA2 (18), ATM (8), BRCA1 (6), CDKN2A (4), PALB2 (3), Lynch (2), and Peutz-Jeghers (1). A family history of PC was present in ≥ 2 FDR in 12 (16%) pts, 1 FDR and 1 SDR in 5 (7) pts, and ≥ 3 ADR in 16 (21%) pts. Of the 65 pts who received screening MRI, 28 (43%) pts had pancreatic cystic lesions identified, including 1 (1%) patient in whom a cholangiocarcinoma was diagnosed as well. No patient underwent surgical resection. Using a 3.0 T MRI to screen patients at high risk for developing PC identified radiographic abnormalities in 43% of patients, which were stable on subsequent surveillance. Specific guidelines for the frequency of surveillance and indications for surgery remain areas of active investigation as the global experience with high risk screening continues to mature.

Prostate cancer is the most common noncutaneous cancer in men in the United States. The current paradigm for screening and diagnosis is imperfect, with relatively low specificity, high cost, and high morbidity. This study aims to generate new image contrasts by learning a distribution of unique image signatures associated with prostate cancer. In total, 48 patients were prospectively recruited for this institutional review board–approved study. Patients underwent multiparametric magnetic resonance imaging 2 weeks before surgery. Postsurgical tissues were annotated by a pathologist and aligned to the in vivo imaging. Radiomic profiles were generated by linearly combining 4 image contrasts (T2, apparent diffusion coefficient [ADC] 0-1000, ADC 50-2000, and dynamic contrast-enhanced) segmented using global thresholds. The distribution of radiomic profiles in high-grade cancer, low-grade cancer, and normal tissues was recorded, and the generated probability values were applied to a naive test set. The resulting Gleason probability maps were stable regardless of training cohort, functioned independent of prostate zone, and outperformed conventional clinical imaging (area under the curve [AUC] = 0.79). Extensive overlap was seen in the most common image signatures associated with high- and low-grade cancer, indicating that low- and high-grade tumors present similarly on conventional imaging.

Gallbladder adenomyomatosis (GA) is a benign gallbladder entity discovered as an asymptomatic gallbladder mass. Since gallbladder cancer is in the differential diagnosis for gallbladder masses, the ability to differentiate benign disease avoids a more extensive oncologic resection. This study sought to review imaging modalities used to diagnose GA.PubMed and SciVerse Scopus were systematically searched using the terms: "gallbladder adenomyomatosis" and "gallbladder imaging" for articles published between January 2000 and January 2015.A total of 14 articles were reviewed in this analysis. Contemporary series report the use of ultrasound (US), computed tomography (CT) or magnetic resonance imaging (MRI) in GA imaging. Ultrasound detection of Rokitansky-Aschoff sinuses, visualized as small cystic spaces with associated "comet-tail" or "twinkling" artifact, is pathognomonic for GA. A "Pearl-Necklace" sign of small connected sinuses on MRI or "Rosary" sign on CT are additional characteristics that may assist in establishing a diagnosis.Ultrasound is the most commonly used tool to investigate GA. If not diagnostic, CT or MRI are effective in attempting to differentiate a benign or malignant cholecystic mass. Characteristic signs should lead the surgeon to perform a laparoscopic cholecystectomy in symptomatic patients or manage non-operatively in asymptomatic patients.

Multiparametric magnetic resonance imaging (MP-MRI), including diffusion-weighted imaging, is commonly used to diagnose prostate cancer. This radiology–pathology study correlates prostate cancer grade and morphology with common b-value combinations for calculating apparent diffusion coefficient (ADC). Thirty-nine patients undergoing radical prostatectomy were recruited for MP-MRI prior to surgery. Diffusion imaging was collected with seven b-values, and ADC was calculated. Excised prostates were sliced in the same orientation as the MRI using 3-D printed slicing jigs. Whole-mount slides were digitized and annotated by a pathologist. Annotated samples were aligned to the MRI, and ADC values were extracted from annotated peripheral zone (PZ) regions. A receiver operating characteristic (ROC) analysis was performed to determine accuracy of tissue type discrimination and optimal ADC b-value combination. ADC significantly discriminates Gleason (G) G4-5 cancer from G3 and other prostate tissue types. The optimal b-values for discriminating high from low-grade and noncancerous tissue in the PZ are 50 and 2000, followed closely by 100 to 2000 and 0 to 2000. Optimal ADC cut-offs are presented for dichotomized discrimination of tissue types according to each b-value combination. Selection of b-values affects the sensitivity and specificity of ADC for discrimination of prostate cancer.

Purpose To examine the effect of low b-values (0

OBJECTIVES/SPECIFIC AIMS: The objective of this study is to use machine Learning techniques to generate maps of epithelium and lumen density in MRI space. METHODS/STUDY POPULATION: Methods: We prospectively recruited 39 patients undergoing prostatectomy for this institutional review board (IRB) approved study. Patients underwent MP-MRI before prostatectomy on a 3T field strength MRI scanner (General Electric, Waukesha, WI, USA) using an endorectal coil. MP-MRI included field-of-view optimized and constrained undistorted single shot (FOCUS) diffusion weighted imaging with 10 b-values (b=0, 10, 25, 50, 80, 100, 200, 500, 1000, and 2000), dynamic contrast enhanced imaging, and T2-weighted imaging. T2 weighted images were intensity normalized and apparent diffusion coefficient maps were calculated. The dynamic contrast enhanced data was used to calculate the percent change in signal intensity before and after contrast injection. All images were aligned to the T2 weighted image. Robotic prostatectomy was performed 2 weeks after image acquisition. Prostate samples were sliced using a 3D printed slicing jig matching the slice profile of the T2 weighted image. Whole mount samples at 10 μm thickness were taken, hematoxylin and eosin stained, digitized, and annotated by a board certified pathologist. A total of 210 slides were included in this study. Lumen and epithelium were automatically segmented using a custom algorithm written in MATLAB. The algorithm was validated by comparing manual to automatic segmentation on 18 samples. Slides were aligned with the T2 weighted image using a nonlinear control point warping technique. Lumen and epithelium density and the expert annotation were subsequently transformed into MRI space. Co-registration was validated by applying a known warp to tumor masks noted by the pathologist and control point warping the whole mount slide to match the transform. Overlap was measured using a DICE coefficient. A learning curve was generated to determine the optimal number of patients to train the algorithm on. A PLS algorithm was trained on 150 random permutations of patients incrementing from 1 to 29 patients. Slides were stratified such that all slides from a single patient were in the same cohort. Three cohorts were generated, with tumor burden balanced across all cohort. A PLS algorithm was trained on 2 independent training sets (cohorts 1 and 2) and applied to cohort 3. The input vector consisted of MRI values and the target variable was lumen and epithelium density. The algorithm was trained lesion-wise. Trained PiCT models were applied to the test cohort voxel-wise to generate 2 new image contrasts. Mean lesion values were compared between high grade, low grade, and healthy tissue using an ANOVA. An ROC analysis was performed lesion-wise on the test set. RESULTS/ANTICIPATED RESULTS: Results: The segmentation accuracy validation revealed R=0.99 and R=0.72 (p

Gadoxetic acid is a liver-specific intravenous T1 magnetic resonance (MR) contrast agent that is excreted via the hepatobiliary system. We hypothesize that hepatocyte expressions of bile transporters (OATP1 and MRP2) correlate with dynamic profile of Gadoxetic acid enhanced (GE)-MR imaging (MRI).Two groups of rats, control (n = 6) and cirrhosis (n = 12), received gadoxetic acid enhanced MRI followed by 70% hepatectomy. The change in MR signal intensity from the baseline before the contrast injection (ΔSI) was analyzed every minute for 30 min. Dynamic signal intensity retention ratio (DSR) was defined as the mean ΔSI of the third 10-minmin period divided by the first 10-minmin period. Real-time PCR was utilized to quantify mRNA expressions.Compared to the control, cirrhosis group demonstrated lower mRNA levels of OATP1 (0.038 ± 0.020 vs. 0.232 ± 0.0979; p = 0.004), MRP2 (0.201 ± 0.084 vs. 0.7567 ± 0.254; p = 0.002), and OATP1/MRP2 mRNA ratio (0.193 ± 0.065 vs. 0.342 ± 0.206; p = 0.032). DSR was higher in the cirrhosis group (0.678 ± 0.554 vs -0.125 ± 0.839; p = 0.033). In the cirrhosis group, there was an inverse correlation between the ratios of OATP1/MRP2 mRNA and DSR (R = -0.709, p = 0.01).Bile transporters OATP1/MRP2 mRNA expression ratio in rat liver tissue decreased with DMN-induced liver injury. The expressions of bile transporters correlated with GE-MRI DSR. The GE-MRI DSR has potential utility in qualifying OATP1/MRP2 mRNA expression.

Objective. Recent data indicate that patients after the Fontan procedure are at risk for significant liver dysfunction; however, the prevalence and extent of liver disease in the Fontan population remains unknown. Furthermore, limited data exist in regard to screening for liver disease in adult Fontan patients. We sought to determine the prevalence of liver disease in adult patients following the Fontan procedure using computed tomography (CT) and serum biomarkers of liver fibrosis. Design. Adult Fontan patients underwent screening for liver disease as part of their annual evaluation. Screening consisted of laboratory evaluation and dual-phase liver CT scan. Laboratory evaluation included analysis of liver function, viral hepatitis serologies, and FibroSURE panel (LabCorp), a test that analyzes the results of serum biomarkers to provide a quantitative surrogate marker for liver fibrosis. Results. Sixteen patients, mean age 30.3 (range 20–41) years, were enrolled in the study. Mean length of follow-up from time of Fontan palliation was 20.5 (range 11–33) years. No patients had serologic evidence of viral hepatitis or synthetic liver dysfunction. Twelve patients (75%) had abnormal FibroSURE scores, seven (44%) had elevated FibroSURE scores predictive of Metavir fibrosis stage F2 or greater on liver biopsy, and one (6%) had a FibroSURE score predictive of cirrhosis on biopsy. All 16 patients had abnormal radiologic liver findings identified on CT, including heterogeneous enhancement in 11 (69%), varices in six (38%), and liver nodules in five patients (31%). Length of time since Fontan surgery correlated significantly with an elevated FibroSURE score (P = .05) and having more CT scan abnormalities (P = .04). Conclusions. Liver fibrosis detected by serum biomarkers and dual phase CT scan is common in adult patients following the Fontan procedure. Further studies are needed to determine the long-term clinical significance of these findings.

Pancreas transplantation aims to restore physiologic normoglycemia in diabetic patients with glomerulopathy and avoid or delay the onset of diabetic retinopathy and arteriopathy. Simultaneous pancreas-kidney transplant is the most common approach, using a cadaveric pancreas donation in conjunction with either cadaveric or live donor renal transplant. Alternative techniques include pancreas after kidney transplant, in which the pancreas transplant is performed some years after renal transplant. Pancreas transplant alone is utilized rarely in diabetic patients with compensated renal function. Pancreas grafts have vascular and enteric connections that vary in their anatomic approach, and understanding of this is critical for imaging with ultrasonography, computed tomography, or magnetic resonance imaging. Imaging techniques are directed to display the pancreatic transplant arterial and venous vasculature, parenchyma, and intestinal drainage pathway. Critical vascular information includes venous thrombosis (partial or complete), arterial occlusion, or aneurysm. Parenchymal abnormalities are nonspecific and occur in pancreatitis, graft rejection, and subsequent graft ischemia. Peripancreatic fluid collections include hematoma/seroma, pseudocyst, and abscess. The latter two are related to pancreatitis, duct disruption, or leak from the duodenojejunostomy. An understanding of transplant anatomy and complications will lead to appropriate use of imaging techniques to diagnose or exclude important complications.

he discovery of pancreatic calcifications has long been used in the diagnosis of pancreatic disease, and for many years imaging of the pancreas was largely limited to the radiographic identification of these calcifications. Today, our ability to image the pancreas has greatly advanced with modalities including CT, sonography, endoscopic retrograde cholangiopancreatography, and MR imaging. This improved visualization of the pancreas allows better identification of pancreatic calcifications and their underlying cause. Traditionally, pancreatic calcifications have been largely associated with chronic calcific pancreatitis from alcohol abuse. Although alcohol abuse remains the predominant cause of pancreatic calcifications, many other causes also deserve attention. Knowing what entities cause calcifications and their typical appearance can help in an accurate diagnosis. However, imaging alone cannot be trusted for a definitive diagnosis because many of these entities overlap in their appearance [1]. We report a spectrum of causes and imaging appearances related to calcification of the pancreas.

Multiparametric magnetic resonance imaging (MP-MRI), including diffusion-weighted imaging, is commonly used to diagnose prostate cancer. This radiology–pathology study correlates prostate cancer grade and morphology with common b-value combinations for calculating apparent diffusion coefficient (ADC). Thirty-nine patients undergoing radical prostatectomy were recruited for MP-MRI prior to surgery. Diffusion imaging was collected with seven b-values, and ADC was calculated. Excised prostates were sliced in the same orientation as the MRI using 3-D printed slicing jigs. Whole-mount slides were digitized and annotated by a pathologist. Annotated samples were aligned to the MRI, and ADC values were extracted from annotated peripheral zone (PZ) regions. A receiver operating characteristic (ROC) analysis was performed to determine accuracy of tissue type discrimination and optimal ADC b-value combination. ADC significantly discriminates Gleason (G) G4-5 cancer from G3 and other prostate tissue types. The optimal b-values for discriminating high from low-grade and noncancerous tissue in the PZ are 50 and 2000, followed closely by 100 to 2000 and 0 to 2000. Optimal ADC cut-offs are presented for dichotomized discrimination of tissue types according to each b-value combination. Selection of b-values affects the sensitivity and specificity of ADC for discrimination of prostate cancer.

Magnetic resonance (MR) angiography is a widely used, noninvasive tool for evaluating the aorta and its branches. It is particularly useful in renal transplant recipients because it provides anatomic detail of the transplant artery without nephrotoxic effects. Volume rendering is underutilized in MR angiography, but this technique affords high-quality three-dimensional MR angiograms, especially in cases of tortuous or complex vascular anatomy. An imaging protocol was developed that includes gadolinium-enhanced MR angiography of the transplant renal artery with volume rendering and multiplanar reformation postprocessing techniques. Axial T2-weighted and contrast material-enhanced T1-weighted MR images are also obtained to examine the renal parenchyma itself and to evaluate for hydronephrosis or peritransplant fluid collections. This imaging protocol allows rapid global assessment of the renal transplant arterial system, renal parenchyma, and peritransplant region. It can also help detect or exclude many of the various causes of renal transplant dysfunction (eg, stenosis or occlusion of a transplant vessel, peritransplant fluid collections, ureteral obstruction). Conventional angiography can thus be avoided in patients with normal findings and reserved for those with MR angiographic evidence of stenosis.

Most cases of abnormal placentation are associated with a history of one or more cesarean deliveries. Uterine leiomyomas and treatment for such a diagnosis are also risk factors for placenta accreta and should be viewed as such.A 34-year-old woman underwent a hysteroscopic myomectomy and became pregnant 6 months later. Ultrasonography and magnetic resonance imaging suggested a placenta percreta. Multidisciplinary care allowed for a safe delivery of her neonate and little maternal morbidity.Patients with a history of hysteroscopic myomectomy or other uterine leiomyoma treatment are at an increased risk for abnormal placentation. Imaging studies are suggested in such patients to coordinate multidisciplinary care to decrease maternal and fetal morbidity and mortality.

Recent data indicate that patients after the Fontan procedure are at risk for significant liver dysfunction; however, the prevalence and extent of liver disease in the Fontan population remains unknown. Furthermore, limited data exist in regard to screening for liver disease in adult Fontan patients. We sought to determine the prevalence of liver disease in adult patients following the Fontan procedure using computed tomography (CT) and serum biomarkers of liver fibrosis. Adult Fontan patients underwent screening for liver disease as part of their annual evaluation. Screening consisted of laboratory evaluation and dual-phase liver CT scan. Laboratory evaluation included analysis of liver function, viral hepatitis serologies, and FibroSURE panel (LabCorp), a test that analyzes the results of serum biomarkers to provide a quantitative surrogate marker for liver fibrosis. Sixteen patients, mean age 30.3 (range 20-41) years, were enrolled in the study. Mean length of follow-up from time of Fontan palliation was 20.5 (range 11-33) years. No patients had serologic evidence of viral hepatitis or synthetic liver dysfunction. Twelve patients (75%) had abnormal FibroSURE scores, seven (44%) had elevated FibroSURE scores predictive of Metavir fibrosis stage F2 or greater on liver biopsy, and one (6%) had a FibroSURE score predictive of cirrhosis on biopsy. All 16 patients had abnormal radiologic liver findings identified on CT, including heterogeneous enhancement in 11 (69%), varices in six (38%), and liver nodules in five patients (31%). Length of time since Fontan surgery correlated significantly with an elevated FibroSURE score (P = .05) and having more CT scan abnormalities (P = .04). Liver fibrosis detected by serum biomarkers and dual phase CT scan is common in adult patients following the Fontan procedure. Further studies are needed to determine the long-term clinical significance of these findings.

The objective of this study is to quantify the contributions of cations, anions and water to stability and specificity of the interaction of lac repressor (lac R) protein with the strong-binding symmetric lac operator (Osym) DNA site. To this end, binding constants Kobs and their power dependences on univalent salt (MX) concentration (SKobs = d log Kobs/d log[MX]) have been determined for the interactions of lac R with Osym operator and with non-operator DNA using filter binding and DNA cellulose chromatography, respectively. For both specific and non-specific binding of lac R, Kobs at fixed salt concentration [KX] increases when chloride (Cl-) is replaced by the physiological anion glutamate (Glu-). At 0.25 M-KX, the increase in Kobs for Osym is observed to be approximately 40-fold, whereas for non-operator DNA the increase in Kobs is estimated by extrapolation to be approximately 300-fold. For non-operator DNA, SKobsRD is independent of salt concentration within experimental uncertainty, and is similar in KCl (SKobs,RDKCl = -9.8(+/- 1.0) between 0.13 M and 0.18 M-KCl) and KGlu (SKobs,RDKGlu = -9.3(+/- 0.7) between 0.23 M and 0.36 M-KGlu). For Osym DNA, SKobsRO varies significantly with the nature of the anion, and, at least in KGlu appears to decrease in magnitude with increasing [KGlu]. Average magnitudes of SKobsRO are less than SKobsRD, and, for specific binding decrease in the order [SKobsRO,KCl[>[SKobsRO,KAc[>[SKobsRO,KGlu[ . Neither KobsRO nor SKobsRO is affected by the choice of univalent cation M+ (Na+, K+, NH4+, or mixtures thereof, all as the chloride salt), and SKobsRO is independent of [MCl] in the range examined (0.125 to 0.3 M). This behavior of SKobsRO is consistent with that expected for a binding process with a large contribution from the polyelectrolyte effect. However, the lack of an effect of the nature of the cation on the magnitude of KobsRO at a fixed [MX] is somewhat unexpected, in view of the order of preference of cations for the immediate vicinity of DNA (NH4+ > K+ > Na+) observed by 23Na nuclear magnetic resonance. For both specific and non-specific binding, the large stoichiometry of cation release from the DNA polyelectrolyte is the dominant contribution to SKobs. To interpret these data, we propose that Glu- is an inert anion, whereas Ac- and Cl- compete with DNA phosphate groups in binding to lac repressor. A thermodynamic estimate of the minimum stoichiometry of water release from lac repressor and Osym operator (210(+/- 30) H2O) is determined from analysis of the apparently significant reduction in [SKobsRO,KGlu[ with increasing [KGlu] in the range 0.25 to 0.9 M. According to this analysis, SKobs values of specific and non-specific binding in KGlu differ primarily because of the release of water in specific binding. In KAc and KCl, we deduce that anion competition affects Kobs and SKobs to an extent which differs for different anions and for the different binding modes.

The aim of this study was to evaluate a new injection-acquisition technique performed using a multirow detector CT scanner for separation of three distinct hepatic circulatory phases (hepatic artery, portal venous inflow, hepatic venous) and to determine which of these phases is optimal for detecting hypervascular neoplasm.Two sequential acquisitions were performed during a single breath-hold followed by a third acquisition beginning 60 sec after injection. Injection-to-scan delay for the first acquisition was the individual patient's circulation time, which was determined by a preliminary mini bolus. The mean attenuation of the upper abdominal aorta, portal vein, and hepatic parenchyma were determined for each imaging pass in 20 patients with cirrhosis and 20 patients without cirrhosis. Tumor-to-liver contrast for hypervascular primary and metastatic neoplasm was evaluated in a different set of 16 cirrhotic patients and nine noncirrhotic patients. Three-dimensional CT arteriograms were obtained from first-pass data.Three distinct circulatory phases (hepatic artery, portal vein inflow or late arterial, and hepatic venous) were seen in cirrhotic and noncirrhotic patients. Maximum tumor-to-liver contrast for hypervascular primary and metastatic neoplasm occurred during the second pass for both cirrhotic (p0.006) and noncirrhotic (p0. 001) patients. A three-dimensional hepatic-mesenteric CT arteriogram of normal or anomalous hepatic vessels without venous overlay was obtained from first-pass data in all patients.Rapid-sequence hepatic helical CT allows selection of the optimal time interval for hypervascular tumor detection. A new paradigm for rapid hepatic CT acquisition-namely, hepatic arterial, portal vein inflow, and hepatic venous phases-is recommended to replace hepatic artery dominant and portal venous phases.