Natural language processing pipeline to extract prostate cancer-related information from clinical notes.

OBJECTIVES: To develop an automated pipeline for extracting prostate cancer-related information from clinical notes. MATERIALS AND METHODS: This retrospective study included 23,225 patients who underwent prostate MRI between 2017 and 2022. Cancer risk factors (family history of cancer and digital rectal exam findings), pre-MRI prostate pathology, and treatment history of prostate cancer were extracted from free-text clinical notes in English as binary or multi-class classification tasks. Any sentence containing pre-defined keywords was extracted from clinical notes within one year before the MRI. After manually creating sentence-level datasets with ground truth, Bidirectional Encoder Representations from Transformers (BERT)-based sentence-level models were fine-tuned using the extracted sentence as input and the category as output. The patient-level output was determined by compilation of multiple sentence-level outputs using tree-based models. Sentence-level classification performance was evaluated using the area under the receiver operating characteristic curve (AUC) on 15% of the sentence-level dataset (sentence-level test set). The patient-level classification performance was evaluated on the patient-level test set created by radiologists by reviewing the clinical notes of 603 patients. Accuracy and sensitivity were compared between the pipeline and radiologists. RESULTS: Sentence-level AUCs were ≥ 0.94. The pipeline showed higher patient-level sensitivity for extracting cancer risk factors (e.g., family history of prostate cancer, 96.5% vs. 77.9%, p < 0.001), but lower accuracy in classifying pre-MRI prostate pathology (92.5% vs. 95.9%, p = 0.002) and treatment history of prostate cancer (95.5% vs. 97.7%, p = 0.03) than radiologists, respectively. CONCLUSION: The proposed pipeline showed promising performance, especially for extracting cancer risk factors from patient's clinical notes. CLINICAL RELEVANCE STATEMENT: The natural language processing pipeline showed a higher sensitivity for extracting prostate cancer risk factors than radiologists and may help efficiently gather relevant text information when interpreting prostate MRI. KEY POINTS: When interpreting prostate MRI, it is necessary to extract prostate cancer-related information from clinical notes. This pipeline extracted the presence of prostate cancer risk factors with higher sensitivity than radiologists. Natural language processing may help radiologists efficiently gather relevant prostate cancer-related text information.

Liver Fibrosis, Fat, and Iron Evaluation with MRI and Fibrosis and Fat Evaluation with US: A Practical Guide for Radiologists.

Quantitative imaging biomarkers of liver disease measured by using MRI and US are emerging as important clinical tools in the management of patients with chronic liver disease (CLD). Because of their high accuracy and noninvasive nature, in many cases, these techniques have replaced liver biopsy for the diagnosis, quantitative staging, and treatment monitoring of patients with CLD. The most commonly evaluated imaging biomarkers are surrogates for liver fibrosis, fat, and iron. MR elastography is now routinely performed to evaluate for liver fibrosis and typically combined with MRI-based liver fat and iron quantification to exclude or grade hepatic steatosis and iron overload, respectively. US elastography is also widely performed to evaluate for liver fibrosis and has the advantage of lower equipment cost and greater availability compared with those of MRI. Emerging US fat quantification methods can be performed along with US elastography. The author group, consisting of members of the Society of Abdominal Radiology (SAR) Liver Fibrosis Disease-Focused Panel (DFP), the SAR Hepatic Iron Overload DFP, and the European Society of Radiology, review the basics of liver fibrosis, fat, and iron quantification with MRI and liver fibrosis and fat quantification with US. The authors cover technical requirements, typical case display, quality control and proper measurement technique and case interpretation guidelines, pitfalls, and confounding factors. The authors aim to provide a practical guide for radiologists interpreting these examinations. © RSNA, 2023 See the invited commentary by Ronot in this issue. Quiz questions for this article are available in the supplemental material.

Phase IB study of sorafenib and evofosfamide in patients with advanced hepatocellular and renal cell carcinomas (NCCTG N1135, Alliance).

Background Sorafenib (Sor) remains a first-line option for hepatocellular carcinoma (HCC) or refractory renal cell carcinomas (RCC). PLC/PRF/5 HCC model showed upregulation of hypoxia with enhanced efficacy when Sor is combined with hypoxia-activated prodrug evofosfamide (Evo). Methods This phase IB 3 + 3 design investigated 3 Evo dose levels (240, 340, 480 mg/m2 on days 8, 15, 22), combined with Sor 200 mg orally twice daily (po bid) on days 1-28 of a 28-day cycle. Primary objectives included determining maximum tolerated dose (MTD) and recommended phase II dose (RP2D) of Sor + Evo. Results Eighteen patients were enrolled (median age 62.5 years; 17 male /1 female; 12 HCC/6 RCC) across three dose levels (DL0: Sor 200 mg bid/Evo 240 mg/m2 [n = 6], DL1:Sor 200 mg bid/Evo 480 mg/m2 [n = 5], DL1a: Sor 200 mg bid/Evo 340 mg/m2 [n = 7]). Two dose-limiting toxicities (DLTs) were reported with Evo 480 mg/m2 (grade 3 mucositis, grade 4 hepatic failure). Grade 3 rash DLT was observed in one patient at Evo 240 mg/m2. No DLTs were observed at Evo 340 mg/m2. MTD and RP2D were established as Sor 200 mg/Evo 340 mg/m2 and Sor 200/Evo 240 mg/m2, respectively. The most common treatment-related adverse events included fatigue, hand-foot syndrome, hypertension, and nausea/vomiting. Two partial responses were observed, one each at DL0 and DL1a.; disease control rate was 55%. Conclusions RP2D was established as sorafenib 200 mg bid + Evo 240 mg/m2. While preliminary anti-tumor activity was observed, future development must account for advances in immunotherapy in HCC/RCC.

Clinical Outcomes of Portosystemic Shunts on the Outcome of Liver Transplantation.

Spontaneous portosystemic shunts (SPSSs) have been associated with worse clinical outcomes in the pre-liver transplantation (LT) setting, but little is known about their post-LT impacts. Our aim was to compare LT candidates with and without SPSSs and assess the impact of SPSSs on patient mortality and graft survival in the post-LT setting. Patients 18 years or older with abdominal imaging done prior to LT were included. Exclusion criteria were the presence of pre-LT surgical shunts, LT indications other than cirrhosis, and combined solid organ transplantations. SPSSs were classified as absent, small, or large according to their maximum diameter (8 mm). Multiple variables that could influence the post-LT course were extracted for analysis. Patient and graft survival were estimated using the Kaplan-Meier method and were compared between groups using a log-rank test. The project received institutional review board approval. We extracted data from 326 patients. After comparing patients without SPSS or with small or large SPSSs, no statistical difference was found for overall patient survival: no SPSS (n = 8/63), reference; small SPSS (n = 18/150), hazard ratio (HR), 1.05 (95% confidence interval [CI], 0.45-2.46); and large SPSS (n = 6/113), HR, 0.60 (95% CI, 0.20-1.78); P = 0.20. Also, no difference was found for graft survival: no SPSS (n = 11/63), reference; small SPSS (n = 21/150), HR, 0.80 (95% CI, 0.38-1.70); large SPSS (n = 11/113), HR, 0.59 (95% CI, 0.25-1.40); P = 0.48. Similarly, no statistical significance was found for these variables when comparing if the graft used was procured from a donation after circulatory death donor versus a donation after brain death donor. In conclusion, the previously described association between SPSSs and worse clinical outcomes in pre-LT patients seems not to persist once patients undergo LT. This study suggests that no steps to correct SPSS intraoperatively are necessary.

Deep Learning Based on MRI for Differentiation of Low- and High-Grade in Low-Stage Renal Cell Carcinoma.

Pretreatment determination of renal cell carcinoma aggressiveness may help to guide clinical decision-making. PURPOSE: To evaluate the efficacy of residual convolutional neural network using routine MRI in differentiating low-grade (grade I-II) from high-grade (grade III-IV) in stage I and II renal cell carcinoma. STUDY TYPE: Retrospective. POPULATION: In all, 376 patients with 430 renal cell carcinoma lesions from 2008-2019 in a multicenter cohort were acquired. The 353 Fuhrman-graded renal cell carcinomas were divided into a training, validation, and test set with a 7:2:1 split. The 77 WHO/ISUP graded renal cell carcinomas were used as a separate WHO/ISUP test set. FIELD STRENGTH/SEQUENCE: 1.5T and 3.0T/T2 -weighted and T1 contrast-enhanced sequences. ASSESSMENT: The accuracy, sensitivity, and specificity of the final model were assessed. The receiver operating characteristic (ROC) curve and precision-recall curve were plotted to measure the performance of the binary classifier. A confusion matrix was drawn to show the true positive, true negative, false positive, and false negative of the model. STATISTICAL TESTS: Mann-Whitney U-test for continuous data and the chi-square test or Fisher's exact test for categorical data were used to compare the difference of clinicopathologic characteristics between the low- and high-grade groups. The adjusted Wald method was used to calculate the 95% confidence interval (CI) of accuracy, sensitivity, and specificity. RESULTS: The final deep-learning model achieved a test accuracy of 0.88 (95% CI: 0.73-0.96), sensitivity of 0.89 (95% CI: 0.74-0.96), and specificity of 0.88 (95% CI: 0.73-0.96) in the Fuhrman test set and a test accuracy of 0.83 (95% CI: 0.73-0.90), sensitivity of 0.92 (95% CI: 0.84-0.97), and specificity of 0.78 (95% CI: 0.68-0.86) in the WHO/ISUP test set. DATA CONCLUSION: Deep learning can noninvasively predict the histological grade of stage I and II renal cell carcinoma using conventional MRI in a multiinstitutional dataset with high accuracy. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2.

MRI - ultrasound fusion guided biopsy of the prostate: lesion volume as a predictor of cancer in patients with repeat biopsies.

INTRODUCTION: The objective was to analyze the diagnostic value of multiparametric magnetic resonance imaging (MRI) prostate lesion volume (PLV) and its correlation with the subsequent MRI-ultrasound (MRI-US) fusion biopsy results. MATERIALS AND METHODS: Between March 2014 and July 2016, 150 men underwent MRI-US fusion biopsies at our institution. All suspicious prostate lesions were graded according to the Prostate Imaging Reporting and Data System (PIRADS) and their volumes were measured. These lesions were subsequently biopsied. All data were prospectively collected and retrospectively analyzed. The PLV of all suspicious lesions was correlated with the presence of cancer on the final MRI-US fusion biopsy. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. RESULTS: There were 206 suspicious lesions identified in 150 men. The overall cancer detection rate was 102/206 (49.5%). The mean PLV for benign lesions was 0.63 ± 0.94 cm3 versus 1.44 ± 1.76 cm3 for cancerous lesions (P < 0.01). There was a statistically significant difference between the PLV of PIRADS 5 lesions when compared to PIRADS 4, 3, and 2 lesions (P < 0.0001, < 0.0001, and 0.006, respectively). The area under the curve for volume in predicting prostate cancer (PCa) was 0.66. The optimal volume for predicting PCa was 0.26 cm3 with a sensitivity, specificity, PPV, and NPV of 80.7%, 42.7%, 41.2%, and 74.6%, respectively. CONCLUSION: PLV may serve as a useful measure to triage patients prior to MRI-US fusion biopsy and help better understand the limits of this technology for individual patients.

Integrated genomic characterization reveals novel, therapeutically relevant drug targets in FGFR and EGFR pathways in sporadic intrahepatic cholangiocarcinoma.

Advanced cholangiocarcinoma continues to harbor a difficult prognosis and therapeutic options have been limited. During the course of a clinical trial of whole genomic sequencing seeking druggable targets, we examined six patients with advanced cholangiocarcinoma. Integrated genome-wide and whole transcriptome sequence analyses were performed on tumors from six patients with advanced, sporadic intrahepatic cholangiocarcinoma (SIC) to identify potential therapeutically actionable events. Among the somatic events captured in our analysis, we uncovered two novel therapeutically relevant genomic contexts that when acted upon, resulted in preliminary evidence of anti-tumor activity. Genome-wide structural analysis of sequence data revealed recurrent translocation events involving the FGFR2 locus in three of six assessed patients. These observations and supporting evidence triggered the use of FGFR inhibitors in these patients. In one example, preliminary anti-tumor activity of pazopanib (in vitro FGFR2 IC50≈350 nM) was noted in a patient with an FGFR2-TACC3 fusion. After progression on pazopanib, the same patient also had stable disease on ponatinib, a pan-FGFR inhibitor (in vitro, FGFR2 IC50≈8 nM). In an independent non-FGFR2 translocation patient, exome and transcriptome analysis revealed an allele specific somatic nonsense mutation (E384X) in ERRFI1, a direct negative regulator of EGFR activation. Rapid and robust disease regression was noted in this ERRFI1 inactivated tumor when treated with erlotinib, an EGFR kinase inhibitor. FGFR2 fusions and ERRFI mutations may represent novel targets in sporadic intrahepatic cholangiocarcinoma and trials should be characterized in larger cohorts of patients with these aberrations.

Diagnostic accuracy of magnetic resonance elastography in liver transplant recipients: A pooled analysis.

BACKGROUND AND AIMS: We conducted an individual participant data (IPD) pooled analysis on the diagnostic accuracy of magnetic resonance elastography (MRE) to detect fibrosis stage in liver transplant recipients. MATERIAL AND METHODS: Through a systematic literature search, we identified studies on diagnostic performance of MRE for staging liver fibrosis, using liver biopsy as gold standard. We contacted study authors for published and unpublished IPD on age, sex, body mass index, liver stiffness, fibrosis stage, degree of inflammation and interval between MRE and biopsy; from these we limited analysis to patients who had undergone liver transplantation. Through pooled analysis using nonparametric two-stage receiver-operating curve (ROC) regression models, we calculated the cluster-adjusted AUROC, sensitivity and specificity of MRE for any (≥ stage 1), significant (≥ stage 2) and advanced fibrosis (≥ stage 3) and cirrhosis (stage 4). RESULTS: We included 6 cohorts (4 published and 2 unpublished series) reporting on 141 liver transplant recipients (mean age, 57 years; 75.2% male; mean BMI, 27.1 kg/m2). Fibrosis stage distribution stage 0, 1, 2, 3, or 4, was 37.6%, 23.4%, 24.8%, 12% and 2.2%, respectively. Mean AUROC values (and 95% confidence intervals) for diagnosis of any (≥ stage 1), significant (≥ stage 2), or advanced fibrosis (≥ stage 3) and cirrhosis were 0.73 (0.66-0.81), 0.69 (0.62-0.74), 0.83 (0.61-0.88) and 0.96 (0.93-0.98), respectively. Similar diagnostic performance was observed in stratified analysis based on sex, obesity and inflammation grade. CONCLUSIONS: In conclusion, MRE has high diagnostic accuracy for detection of advanced fibrosis and cirrhosis in liver transplant recipients, independent of BMI and degree of inflammation.

Relationship between sarcopenia, six-minute walk distance and health-related quality of life in liver transplant candidates.

Sarcopenia, or loss of skeletal muscle mass, is associated with increased mortality and morbidity in liver transplant (LT) candidates. Six-minute walk distance (6MWD) and health-related quality of life (HRQOL) as assessed by short form 36 scores (SF-36) also impact clinical outcomes in these patients. This study explored the relationship between the sarcopenia, 6MWD, and HRQOL in LT candidates. Sarcopenia was evaluated based on skeletal muscle mass index (SMI) quantified from abdominal computed tomography. Patients were followed until death, removal from the wait list or the end of the study period. Two hundred and thirteen patients listed for LT were included. The mean SMI, 6MWD and mean gait speed were 54.3 ± 9.7, 370.5 m and 1 m/s, respectively. Sarcopenia was noted in 22.2% of LT candidates. There was no correlation between sarcopenia, 6MWD, and SF-36 scores. The 6MWD, but not sarcopenia, was an independent predictor of mortality (hazard ratio = 2.1 [0.9-4.7]). In summary, sarcopenia did not emerge as a significant predictor of waitlist mortality and also failed to correlate with either functional capacity or HRQOL in LT candidates. In patients with ESLD awaiting LT, 6MWD appears to be a more useful prognostic indicator than the presence of sarcopenia.

Magnetic resonance elastography (MRE) in assessing hepatic fibrosis: performance in a cohort of patients with histological data.

PURPOSE: The purpose of this study was to compare fibrosis seen on liver biopsy to MR elastography (MRE) stiffness measurements in normal controls and patients with abnormal transaminases and chronic liver disease. METHODS: The control group consisted of 22 healthy liver transplant donors who by definition had normal transaminases. The patient group (32 patients) was recruited from the Mayo Clinic Arizona hepatobiliary clinic over a 3-year span. All subjects underwent a liver biopsy as part of their evaluation and agreed to MRE within 35 days of biopsy. Non-parametric tests were used to compare the MRE-predicted liver fibrosis to the fibrosis noted on liver biopsy. RESULTS: Analysis included 54 subjects (32 patients with chronic liver disease and 22 healthy liver donor candidates). MRE median liver stiffness measurements increased per histologic liver fibrosis stage (one-way ANOVA p = 0.001), with significant correlation between increasing fibrosis stage and stiffness values. Median MRE for control group (2.13 kPa; mean = 2.3 ± 0.6 kPa) was significantly lower than for patient group (3.7 kPa; mean = 4.1 ± 2.1 kPa) (p = 0.003). Significantly, lower median stiffness was seen in zero-to-moderate (F0-2, n = 22) vs. severe fibrosis stages (F3-4, n = 10) 2.80 vs. 5.9 kPa, respectively (p < 0.05). Using a 3.7-kPa cut-off value, the predicted sensitivity and specificity for detecting F0-2 from F3-4 were 91% and 80%, respectively. CONCLUSIONS: Our analysis supports previous findings that MRE is a non-invasive and effective method for detection and assessment of liver fibrosis, particularly for discrimination between F0-2 stages and F3-4 stages. MRE may represent a valuable tool to finely discern hepatic fibrosis non-invasively.

MRI of the stomach: a pictorial review with a focus on oncological applications and gastric motility.

The purpose of this pictorial review is to demonstrate gastric pathology seen on magnetic resonance imaging (MRI) and discuss the essential MRI sequences for the evaluation of benign and malignant gastric pathologies. Common tumors of the stomach, polyposis syndromes, iatrogenic conditions, as well as other conditions of the stomach will be reviewed. The utility of MRI in the evaluation of patients with gastric malignancies and disorders of gastric motility will also be discussed.

Magnetic resonance elastography: evaluation of new inversion algorithm and quantitative analysis method.

PURPOSE: To assess mean shear hepatic stiffness calculations using various region of interest (ROI) techniques, a new inversion algorithm, and a confidence threshold mask. METHODS: Seventy-three patients (49 with abnormal liver function tests/known chronic liver disease and 24 healthy liver transplant donors) underwent liver biopsy and magnetic resonance elastography (MRE). MRE data processed with the current inversion algorithm [multiscale direct inversion (MSDI)] was assessed using 2 ROI methods (single vs. triple). The data were then reprocessed using the new inversion algorithm (multimodel direct inversion [MMDI]) Hepatic stiffness calculations were performed using a single (70%) ROI method, with/without a 95% confidence threshold mask, and compared with MSDI. RESULTS: For MSDI, average stiffness difference between single and triple ROI methods was not statistically significant by the 2-sample t test [0.15 kilopascals (kPa); P = .77]. For the 2 algorithms, there was little difference in average stiffness measurements of MSDI and MMDI (mean, 0.32 kPa; 9%) using a confidence mask with good agreement [intraclass correlation coefficient (ICC), 0.986 (95% CI 0.975-0.994)]. Use of the confidence mask showed excellent consistency and less variance [ICC, 0.995 (95% CI 0.993-0.998)] compared to either the inter-observer or intra-observer freehand technique. CONCLUSION: MRE analysis showed no significant difference between the 2 freehand ROI techniques. With a 9% average kPa variance, stiffness measurements for MSDI and MMDI were also not significantly different. The use of the confidence mask reduces calculated stiffness variability, which impacts the use of MRE for assessing therapy response and initial/longitudinal assessment of chronic liver disease.

Observer performance for adaptive, image-based denoising and filtered back projection compared to scanner-based iterative reconstruction for lower dose CT enterography.

PURPOSE: The purpose of this study was to compare observer performance for detection of intestinal inflammation for low-dose CT enterography (LD-CTE) using scanner-based iterative reconstruction (IR) vs. vendor-independent, adaptive image-based noise reduction (ANLM) or filtered back projection (FBP). METHODS: Sixty-two LD-CTE exams were performed. LD-CTE images were reconstructed using IR, ANLM, and FBP. Three readers, blinded to image type, marked intestinal inflammation directly on patient images using a specialized workstation over three sessions, interpreting one image type/patient/session. Reference standard was created by a gastroenterologist and radiologist, who reviewed all available data including dismissal Gastroenterology records, and who marked all inflamed bowel segments on the same workstation. Reader and reference localizations were then compared. Non-inferiority was tested using Jackknife free-response ROC (JAFROC) figures of merit (FOM) for ANLM and FBP compared to IR. Patient-level analyses for the presence or absence of inflammation were also conducted. RESULTS: There were 46 inflamed bowel segments in 24/62 patients (CTDIvol interquartile range 6.9-10.1 mGy). JAFROC FOM for ANLM and FBP were 0.84 (95% CI 0.75-0.92) and 0.84 (95% CI 0.75-0.92), and were statistically non-inferior to IR (FOM 0.84; 95% CI 0.76-0.93). Patient-level pooled confidence intervals for sensitivity widely overlapped, as did specificities. Image quality was rated as better with IR and AMLM compared to FBP (p < 0.0001), with no difference in reading times (p = 0.89). CONCLUSIONS: Vendor-independent adaptive image-based noise reduction and FBP provided observer performance that was non-inferior to scanner-based IR methods. Adaptive image-based noise reduction maintained or improved upon image quality ratings compared to FBP when performing CTE at lower dose levels.

Magnetic resonance elastography can discriminate normal vs. abnormal liver biopsy in candidates for live liver donation.

PURPOSE: The aim of this study was to define liver shear stiffness by magnetic resonance elastography (MRE) that distinguishes normal from abnormal liver biopsy, especially when steatosis ≥20%, among potential live liver donors. METHODS: Baseline clinical, laboratory, imaging, MRE, and liver biopsy results were recorded. Using MRE, hepatic shear stiffness in kilopascals (kPa) was measured and compared to liver biopsy. Comparison between groups was done using χ(2) or Fisher's exact test for categorical variables and Wilcoxon test for continuous variables. Receiver operating characteristic (ROC) curve was calculated to assess diagnostic accuracy. Statistical significance was set at p < 0.05. RESULTS: 38 healthy adults were included. Liver biopsy was normal in 27 and abnormal in 11. ROC curve for MRE defined optimal cutoff at 2.6 kPa (sensitivity 0.72, specificity 0.85, AUC 0.81) to distinguish these 2 groups. Hepatic steatosis ≥20% on biopsy is a contraindication for liver donation in our center. We evaluated the ability of MRE to distinguish this degree of steatosis: 8 persons had steatosis ≥20% and were excluded from donation. ROC curve for MRE defined optimal cutoff at 2.82 kPa (sensitivity 0.88, specificity 1, AUC 0.98) to identify this group. CONCLUSIONS: Liver stiffness measured by MRE, even in the absence of liver fibrosis, can be useful in differentiating normal from abnormal liver histology, and most importantly in patients under evaluation for live liver donation, can very accurately distinguish those with complicated hepatic steatosis ≥20%, our cutoff for donation. In the future, MRE might provide supplementary information to make liver biopsy unnecessary in the donor evaluation process.

Pilot study: Evaluation of dual-energy computed tomography measurement strategies for positron emission tomography correlation in pancreatic adenocarcinoma.

We sought to determine whether dual-energy computed tomography (DECT) measurements correlate with positron emission tomography (PET) standardized uptake values (SUVs) in pancreatic adenocarcinoma, and to determine the optimal DECT imaging variables and modeling strategy to produce the highest correlation with maximum SUV (SUVmax). We reviewed 25 patients with unresectable pancreatic adenocarcinoma seen at Mayo Clinic, Scottsdale, Arizona, who had PET-computed tomography (PET/CT) and enhanced DECT performed the same week between March 25, 2010 and December 9, 2011. For each examination, DECT measurements were taken using one of three methods: (1) average values of three tumor regions of interest (ROIs) (method 1); (2) one ROI in the area of highest subjective DECT enhancement (method 2); and (3) one ROI in the area corresponding to PET SUVmax (method 3). There were 133 DECT variables using method 1, and 89 using the other methods. Univariate and multivariate analysis regression models were used to identify important correlations between DECT variables and PET SUVmax. Both R2 and adjusted R2 were calculated for the multivariate model to compensate for the increased number of predictors. The average SUVmax was 5 (range, 1.8-12.0). Multivariate analysis of DECT imaging variables outperformed univariate analysis (r = 0.91; R2 = 0.82; adjusted R2 = 0.75 vs. r < 0.58; adjusted R2 < 0.34). Method 3 had the highest correlation with PET SUVmax (R2 = 0.82), followed by method 1 (R2 = 0.79) and method 2 R2 = 0.57). DECT thus has clinical potential as a surrogate for, or as a complement to, PET in patients with pancreatic adenocarcinoma.

Noncathartic CT colonography: Image quality assessment and performance and in a screening cohort.

OBJECTIVE: Cathartic bowel preparation is a major barrier for colorectal cancer screening. We examined noncathartic CT colonography (CTC) quality and performance using four similar bowel-tagging regimens in an asymptomatic screening cohort. SUBJECTS AND METHODS: This prospective study included 564 asymptomatic subjects who underwent noncathartic CTC without dietary modification but with 21 g of barium with or without iodinated oral contrast material (four regimens). The quality of tagging with oral agents was evaluated. A gastrointestinal radiologist evaluated examinations using primary 2D search supplemented by electronic cleansing (EC) and 3D problem solving. Results were compared with complete colonoscopy findings after bowel purgation and with retrospective unblinded evaluation in 556 of the 564 (99%) subjects. RESULTS: Of the 556 subjects, 7% (37/556) and 3% (16/556) of patients had 52 and 20 adenomatous polyps ≥ 6 and ≥ 10 mm, respectively. The addition of iodine significantly improved the percentage of labeled stool (p ≤ 0.0002) and specificity (80% vs 89-93%, respectively; p = 0.046). The overall sensitivity of noncathartic CTC for adenomatous polyps ≥ 6 mm was 76% (28/37; 95% CI, 59-88%), which is similar to the sensitivity of the iodinated regimens with most patients (sensitivity: 231 patients, 74% [14/19; 95% CI, 49-91%]; 229 patients, 80% [12/15; 95% CI, 52-96%]). The negative predictive value was 98% (481/490), and the lone cancer was detected (0.2%, 1/556). EC was thought to improve conspicuity of 10 of 21 visible polyps ≥ 10 mm. CONCLUSION: In this prospective study of asymptomatic subjects, the per-patient sensitivity of noncathartic CTC for detecting adenomas ≥ 6 mm was approximately 76%. Inclusion of oral iodine contrast material improves examination specificity and the percentage of labeled stool. EC may improve polyp conspicuity.

Reducing radiation dose in CT enterography.

Computed tomographic (CT) enterography is a diagnostic examination that is increasingly being used to evaluate disorders of the small bowel. An undesirable consequence of CT, however, is patient exposure to ionizing radiation. This is of particular concern with CT enterography because patients tend to be young and require numerous follow-up examinations. There are multiple strategies to reduce radiation dose at CT enterography, including adjusting acquisition parameters, reducing scan length, and reducing tube voltage or tube current. The drawback to dose reduction strategies is degradation of image quality due to increased image noise. However, image noise can be reduced with commercial iterative reconstruction and denoising techniques. With a combination of low-dose techniques and noise-control strategies, one can markedly reduce radiation dose at CT enterography while maintaining diagnostic accuracy.

Performing Automatic Identification and Staging of Urothelial Carcinoma in Bladder Cancer Patients Using a Hybrid Deep-Machine Learning Approach.

Accurate clinical staging of bladder cancer aids in optimizing the process of clinical decision-making, thereby tailoring the effective treatment and management of patients. While several radiomics approaches have been developed to facilitate the process of clinical diagnosis and staging of bladder cancer using grayscale computed tomography (CT) scans, the performances of these models have been low, with little validation and no clear consensus on specific imaging signatures. We propose a hybrid framework comprising pre-trained deep neural networks for feature extraction, in combination with statistical machine learning techniques for classification, which is capable of performing the following classification tasks: (1) bladder cancer tissue vs. normal tissue, (2) muscle-invasive bladder cancer (MIBC) vs. non-muscle-invasive bladder cancer (NMIBC), and (3) post-treatment changes (PTC) vs. MIBC.

Toward biphasic computed tomography (CT) enteric contrast: material classification of luminal bismuth and mural iodine in a small-bowel phantom using dual-energy CT.

OBJECTIVE: To develop dual-energy computed tomography methods for identification of hyperenhancing, hypoenhancing, and nonenhancing small-bowel pathologies. METHODS: Small-bowel phantoms simulating varying patient sizes and polyp types (hyperenhancing, hypoenhancing, and nonenhancing) contained bismuth suspension in the lumen. Dual-energy CT was performed at 80/140 kV and 100/140 kV. Computed tomographic number ratios (CT numbers at low/high kilovoltage) were calculated. Two radiologists evaluated polyp detection and conspicuity using bismuth-only, iodine-only, iodine-overlay, and mixed-kilovoltage displays. RESULTS: Computed tomographic ratios for bismuth and iodine did not overlap. For hyperenhancing and nonenhancing polyps at 80/140 kV, iodine-overlay display yielded higher detection rate (96%, 94%) and conspicuity score (3.5, 3.1) than mixed-kilovoltage images (88%, 68%; 1.5, 2.7). Mixed-kV images performed slightly better for hypoenhancing polyps (92%, 3.4 vs. <80%, <2.9). Similar results were observed at 100/140kV. CONCLUSIONS: Dual-energy CT and a bismuth-containing enteric contrast permitted simultaneous identification of hyperenhancing, hypoenhancing, and nonenhancing polyps over a range of patient sizes.

Long-term outcomes of prostate intensity-modulated radiation therapy incorporating a simultaneous intra-prostatic MRI-directed boost.

Purpose/objectives: This retrospective study demonstrates the long-term outcomes of treating prostate cancer using intensity modulated (IMRT) with incorporation of MRI-directed boost. Materials/methods: From February 2009 to February 2013, 78 men received image-guided IMRT delivering 77.4 Gy in 44 fractions with simultaneously integrated boost to 81-83 Gy to an MRI-identified lesion. Patients with intermediate-risk or high-risk prostate cancer were recommended to receive 6 and 24-36 months of adjuvant hormonal therapy, respectively. Results: Median follow-up was 113 months (11-147). There were 18 low-risk, 43 intermediate-risk, and 17 high-risk patients per NCCN risk stratification included in this study. Adjuvant hormonal therapy was utilized in 32 patients (41%). The 10-year biochemical control rate for all patients was 77%. The 10-year biochemical control rates for low-risk, intermediate-risk, and high-risk diseases were 94%, 81%, and 88%, respectively (p = 0.35). The 10-year rates of local control, distant control, and survival were 99%, 88%, and 66%, respectively. Of 25 patients who died, only four (5%) died of prostate cancer. On univariate analysis, T-category and pretreatment PSA level were associated with distant failure rate (p = 0.02). There was no grade =3 genitourinary and gastrointestinal toxicities that persisted at the last follow-up. Conclusions: This study demonstrated the long-term efficacy of using MRI to define an intra-prostatic lesion for SIB to 81-83Gy while treating the entire prostate gland to 77.4 Gy with IMRT. Our study confirms that modern MRI can be used to locally intensify dose to prostate tumors providing high long-term disease control while maintaining favorable long-term toxicity.