Flow-Based Spatio-Temporal Structured Prediction of Motion Dynamics.

Conditional Normalizing Flows (CNFs) are flexible generative models capable of representing complicated distributions with high dimensionality and large interdimensional correlations, making them appealing for structured output learning. Their effectiveness in modelling multivariates spatio-temporal structured data has yet to be completely investigated. We propose MotionFlow as a novel normalizing flows approach that autoregressively conditions the output distributions on the spatio-temporal input features. It combines deterministic and stochastic representations with CNFs to create a probabilistic neural generative approach that can model the variability seen in high-dimensional structured spatio-temporal data. We specifically propose to use conditional priors to factorize the latent space for the time dependent modeling. We also exploit the use of masked convolutions as autoregressive conditionals in CNFs. As a result, our method is able to define arbitrarily expressive output probability distributions under temporal dynamics in multivariate prediction tasks. We apply our method to different tasks, including trajectory prediction, motion prediction, time series forecasting, and binary segmentation, and demonstrate that our model is able to leverage normalizing flows to learn complicated time dependent conditional distributions.

Maintaining low non-neoplastic polypectomy rates in high-quality screening colonoscopy.

BACKGROUND AND AIMS: Non-neoplastic polypectomies (NNPs) add pathology and procedural costs but do not reduce cancer risk and should be minimized. We sought to define the minimal non-neoplastic polypectomy rate (NNPR) for those colonoscopists achieving high-quality colorectal cancer screening based on adenoma detection rates (ADRs). METHODS: NNPRs for colonoscopists achieving high-quality adenoma detection rates were reported to determine minimal NNPR goals. Two approaches to tracking NNPR monitoring were compared: (1) total NNPR, an NNPR inclusive of all non-neoplastic specimens with exclusion of only hyperplastic polyp, sessile serrated polyp, and adenoma; and (2) normal tissue-only NNPR, an NNPR inclusive of those specimens with only normal colonic mucosa or lymphoid follicles. RESULTS: For those performing colonoscopy with high-quality ADRs (≥25%), half (6/12) of the colonoscopists had a total NNPR of ≤8.5% and 2 gastroenterologists had a total NNPR of ≤3.4%. The mean total NNPR of the cohort was 8.7% versus the normal tissue only NNPR, which was 7.5% (mean difference of 1.2%, standard deviation ± 0.97). The widest variation between total NNPR versus normal tissue only NNPR for any colonoscopist was 2.9%. The total NNPR ranged between 2.6% and 21.3% among 14 colonoscopists. CONCLUSIONS: Colonoscopy with a high-quality ADR can be achieved while maintaining a low total NNPR. A total NNPR, inclusive of all non-neoplastic specimens as an alternative to an approach in which all specimens require individual review in order to select out only normal tissue can be considered for monitoring of NNPR.

Most Premature Surveillance Colonoscopy Is Not Attributable to Bowel Preparation or New Clinical Indications.

BACKGROUND AND AIMS: Surveillance colonoscopy frequently occurs prior to recommended intervals. Studies delineating the reasons why premature surveillance occurs are limited. We sought to define the frequency in which premature surveillance colonoscopy occurs in the setting of an inadequate bowel preparation or with a provided patient clinical indication versus when premature surveillance colonoscopy occurs without any provided discernible rationale in the setting of adequate bowel preparation. METHODS: A retrospective cross-sectional cohort study of 700 patients undergoing colonoscopy for an indication of "surveillance of polyps" from 2008 to 2014 at two tertiary-care referral centers was carried out. Patients were deemed either "adherent" or "premature" based on US Multi-Society Task Force guideline intervals for surveillance colonoscopy. A documented decision-making rationale for premature surveillance was determined through review of the electronic medical record with assessment of clinical notes and endoscopy order and report. RESULTS: Premature surveillance occurred in 43.0 % (n = 301) of all surveillance colonoscopies performed. Among the premature cases, rationale was attributed to inadequate bowel preparation in 17.3 % (n = 52) and due to a new clinical indication in 21.6 % (n = 65). Most commonly, in 61.1 % (n = 184) of premature cases, no rationale was documented for the early colonoscopy. CONCLUSIONS: Documented decision-making rationale for premature surveillance colonoscopy is usually absent in premature cases with inadequate bowel preparation and new clinical indications explaining only a minority of the occurrences.

Presence of small sessile serrated polyps increases rate of advanced neoplasia upon surveillance compared with isolated low-risk tubular adenomas.

BACKGROUND AND AIMS: The U.S. Multi-Society Task Force (USMSTF) stratifies patients with sessile serrated polyps (SSPs) without cytologic dysplasia of <10 mm in size as at low risk for metachronous advanced neoplasia and recommends management similar to low-risk conventional tubular adenomas. Evidence supporting the recommended surveillance interval for these low-risk SSPs is limited. We aimed to assess rates of metachronous advanced neoplasia based on the presence of an initial low-risk SSP compared with isolated low-risk tubular adenomas. METHODS: Colonoscopy data were retrieved for 2260 patients found to have an adenoma or SSP on pathology records between 2005 and 2011 at an academic medical center. The 788 patients who met study design criteria were stratified into 4 groups based on the presence of a high- or low-risk adenoma (HRA or LRA) and of a synchronous SSP on initial colonoscopy. The rates of advanced neoplasia at surveillance colonoscopy were then compared between groups. RESULTS: The rate of advanced neoplasia at surveillance in the LRA inclusive of SSP group (12/66, 18.2%) was greater than in the LRA without any SSP group (29/370, 7.8%; P = .019). The rate of advanced neoplasia at surveillance in patients with isolated low-risk SSP (10/56, 17.9%) remained significantly greater than those with isolated low-risk tubular adenomas (29/370, 7.8%; P = .024). The rate of advanced neoplasia upon surveillance in the LRA inclusive of SSP group (18.2%) was comparable with the rate observed in the index HRA without any SSP group (15.9%) (40/252, P = .709). CONCLUSIONS: The rate of advanced neoplasia upon surveillance in patients with initial low-risk SSPs is higher than in patients with initial isolated low-risk tubular adenomas and more similar to patients with initial high-risk tubular adenomas. These findings suggest that the rate of metachronous advanced neoplasia in patients with what are considered by USMSTF as "low-risk" SSPs is higher than in those without SSPs. Therefore, a surveillance interval that accounts for the presence of SSPs even in small lesions without cytologic dysplasia should be considered.

Capillary hemangioma of the testis: A rare benign tumour.

Testicular capillary hemangioma is a rare benign vascular tumour. We report a case of a 66-year-old man who underwent an uncomplicated radical orchiectomy for a painless left testicular mass. Pathology showed capillary hemangioma of the testis. There are only 22 cases reported in the English literature, including the presented case. Appropriate intra-operative recognition of this entity is vital to assess for potential testicular-sparing surgery.

Patient Non-adherence and Cancellations Are Higher for Screening Colonoscopy Compared with Surveillance Colonoscopy.

BACKGROUND: A significant proportion of the eligible population is non-adherent to colonoscopy for colorectal cancer (CRC) screening. AIMS: To define the demographic and clinical variables associated with non-adherence and multiple cancellations to scheduled colonoscopy within 1 year in a CRC screening and adenomatous polyp surveillance population. METHODS: This was an observational cohort study of 617 consecutive patients scheduled to undergo colonoscopy at an outpatient academic tertiary care center for CRC screening or adenomatous polyp surveillance from January 2012 to September 2012. RESULTS: Overall, 551 patients (89.3%) were adherent and 66 (10.7%) were non-adherent to scheduled colonoscopy at 1 year. The relative risk for non-adherence was 5.42 [95% confidence interval (CI) 2.74-10.75] in patients undergoing colonoscopy for screening compared to those for surveillance (16.7 vs. 3.5% non-adherence, respectively, P < 0.001). An indication of screening in comparison with surveillance was associated with non-adherence [odds ratio (OR) 12.69, 95% CI 4.18-38.51] and multiple cancellations (OR 2.33, 95% CI 1.27-4.31) by multiple regression analysis. CONCLUSIONS: Patients undergoing colonoscopy for CRC screening are significantly less likely to attend their scheduled procedure within a year and have more procedure cancellations than those undergoing surveillance colonoscopy.

Is there a role for anterior zone sampling as part of saturation trans-rectal ultrasound guided prostate biopsy?

BACKGROUND: The prostatic anterior zone (AZ) is not targeted routinely by TRUS guided prostate biopsy (TRUS-Pbx). MRI is an accurate diagnostic tool for AZ tumors, but is often unavailable due to cost or system restrictions. We examined the diagnostic yield of office based AZ TRUS-Pbx. METHODS: 127 men at risk for AZ tumors were studied: Patients with elevated PSA and previous extended negative TRUS-Pbx (group 1, n = 78) and actively surveyed low risk prostate cancer patients (group 2, n = 49). None of the participants had a previous AZ biopsy. Biopsy template included suspicious ultrasonic areas, 16 peripheral zone (PZ), 4 transitional zone (TZ) and 6 AZ cores. All biopsies were performed by a single urologist under local peri-prostatic anaesthetic, using the B-K Medical US System, an end-firing probe 4-12 MHZ and 18 ga/25 cm needle. All samples were reviewed by a single specialized uro-pathologist. Multivariate analysis was used to detect predictors for AZ tumors accounting for age, PSA, PSA density, prostate volume, BMI, and number of previous biopsies. RESULTS: Median PSA was 10.4 (group 1) and 7.3 (group 2). Age (63.9, 64.5), number of previous biopsies (1.5) and cores (17.8, 21.3) and prostate volume (56.4 cc, 51 cc) were similar for both groups. The overall diagnostic yield was 34.6% (group 1) and 85.7% (group 2). AZ cancers were detected in 21.8% (group 1) and 34.7% (group 2) but were rarely the only zone involved (1.3% and 4.1% respectively). Gleason ≥ 7 AZ cancers were often accompanied by equal grade PZ tumors. In multivariate analysis only prostate volume predicted for AZ tumors. Patients detected with AZ tumors had significantly smaller prostates (36.9 cc vs. 61.1 cc p < 0.001). Suspicious AZ ultrasonic findings were uncommon (6.3%). CONCLUSIONS: TRUS-Pbx AZ sampling rarely improves the diagnostic yield of extended PZ sampling in patients with elevated PSA and previous negative biopsies. In low risk prostate cancer patients who are followed by active surveillance, AZ sampling changes risk stratification in 6% but larger studies are needed to define the role of AZ sampling in this population and its correlation with prostatectomy final pathological specimens.

Commonality and differences of methylation signatures in the plasma of patients with pancreatic cancer and colorectal cancer.

Profiling of DNA methylation status of specific genes is a way to screen for colorectal cancer (CRC) and pancreatic cancer (PC) in blood. The commonality of methylation status of cancer-related tumor suppressor genes between CRC and PC is largely unknown. Methylation status of 56 cancer-related genes was compared in plasma of patients in the following cohorts: CRC, PC and healthy controls. Cross validation determined the best model by area under ROC curve (AUC) to differentiate cancer methylation profiles from controls. Optimal preferential gene methylation signatures were derived to differentiate either cancer (CRC or PC) from controls. For CRC alone, a three gene signature (CYCD2, HIC and VHL) had an AUC 0.9310, sensitivity (Sens) = 0.826, specificity (Spec) = 0.9383. For PC alone, an optimal signature consisted of five genes (VHL, MYF3, TMS, GPC3 and SRBC), AUC 0.848; Sens = 0.807, Spec = 0.666. Combined PC and CRC signature or "combined cancer signature" was derived to differentiate either CRC and PC from controls (MDR1, SRBC, VHL, MUC2, RB1, SYK and GPC3) AUC = 0.8177, Sens = 0.6316 Spec = 0.840. In a validation cohort, N = 10 CRC patients, the optimal CRC signature (CYCD2, HIC and VHL) had AUC 0.900. In all derived signatures (CRC, PC and combined cancer signature) the optimal panel used preferential VHL methylation. In conclusion, CRC and PC differ in specific genes methylated in plasma other than VHL. Preferential methylation of VHL is shared in the optimal signature for CRC alone, PC alone and combined PC and CRC. Future investigations may identify additional methylation markers informative for the presence of both CRC and PC.

Advanced adenoma detection rate is independent of nonadvanced adenoma detection rate.

OBJECTIVES: Adenoma detection rate (ADR) is the accepted rate marker in colonoscopy quality. Advanced adenomas detected at index colonoscopy, while less frequent than nonadvanced adenomas, carry greater risk for future advanced neoplasia during surveillance colonoscopy. This study aimed to determine the effect of the colonoscopist and other factors on advanced ADR and to define the correlation of advanced and nonadvanced ADRs among colonoscopists. METHODS: An observational study of a cohort of patients undergoing first-time colorectal cancer screening colonoscopy was conducted. Patient characteristics and colonoscopic findings were collected. Adenoma, advanced adenoma, and nonadvanced ADRs were calculated. Logistic regression was used to determine variable effects on advanced adenoma detection, and Spearman's rank-order correlation was used to evaluate the relationship between advanced and nonadvanced ADRs. RESULTS: A total of 1,944 patients had first-time screening colonoscopies by 14 colonoscopists. All colonoscopists had adequate (>20%) ADRs. The variability in the colonoscopist ranges of detection was 22.22 to 44.66% for adenomas and 2.00 to 18.18% for advanced adenomas. Logistic regression showed that increasing patient age (odds ratio (OR) 1.16 per 5-year increase, 95% confidence interval (CI) 1.05-1.28, P=0.008) and male gender (OR 2.15, 95% CI 1.51-3.06, P<0.0001) were variables associated with advanced adenoma detection. Colonoscopists were significantly different in detecting advanced adenomas by random effects model (P=0.002), adjusting for patient age, gender, race, year of colonoscopy, gastroenterology fellow participation during colonoscopy, and nonadvanced adenomas. Spearman's rank-order correlation coefficient of -0.42 (95% CI -0.77 to 0.14, P=0.13) was not significant and showed no correlation between advanced and nonadvanced adenoma detection by the group of colonoscopists. CONCLUSIONS: Advanced ADR is variable among colonoscopists with acceptable ADRs. Colonoscopists' advanced ADRs are independent of their nonadvanced ADRs.

Erectile dysfunction and testosterone deficiency syndrome: the "portal to men's health".

Erectile dysfunction (ED) and testosterone deficiency syndrome (TDS) are closely related. In addition to affecting men's sexual health, both conditions also affect other male health issues. Screening for ED, especially in younger men, should become standard clinical practice for the primary care physician. Possible systemic effects and associated effects of TDS are now well documented. Testosterone replacement therapy (TRT) is very safe and effective in the right man.

Commentary: Competency restoration research--complicating an already complex process.

Predicting restorability in individuals found not competent to stand trial is an enduring focus of interest among forensic clinicians and academicians. In our commentary, we suggest that to understand this area even more comprehensively, we must look further. We must build on existing research on fitness to stand trial, move beyond diagnosis and a binary competence variable, and include the complex interplay between symptoms and fitness-related capacities that may be associated with lack of adjudicative competence and challenges to restorability.

The N terminus of connexin37 contains an alpha-helix that is required for channel function.

The cytoplasmic N-terminal domain of connexins has been implicated in multiple aspects of gap junction function, including connexin trafficking/assembly and channel gating. A synthetic peptide corresponding to the first 23 amino acids of human connexin37 was prepared, and circular dichroism and nuclear magnetic resonance studies showed that this N-terminal peptide was predominantly alpha-helical between glycine 5 and glutamate 16. The importance of this structure for localization of the protein at appositional membranes and channel function was tested by expression of site-directed mutants of connexin37 in which amino acids leucine 10 and glutamine 15 were replaced with prolines or alanines. Wild type connexin37 and both substitution mutants localized to appositional membranes between transfected HeLa cells. The proline mutant did not allow intercellular transfer of microinjected neurobiotin; the alanine mutant allowed transfer, but less extensively than wild type connexin37. When expressed alone in Xenopus oocytes, wild type connexin37 produced hemichannel currents, but neither of the double substitution mutants produced detectable currents. The proline mutant (but not the alanine mutant) inhibited co-expressed wild type connexin37. Taken together, our data suggest that the alpha-helical structure of the connexin37 N terminus may be dispensable for protein localization, but it is required for channel and hemichannel function.

Association of troponin status with guideline-based management of acute myocardial infarction in older persons.

BACKGROUND: Over the past decade, a large body of evidence has emerged demonstrating the prognostic significance of troponin as well as its use in tailoring therapeutic interventions. Little is known, however, regarding the association of troponin status with guideline-based therapies in older patients with acute myocardial infarction (AMI). METHODS: A nationwide sample of eligible Medicare beneficiaries 65 years or older, who were hospitalized with a primary discharge diagnosis of AMI from April 1998 to March 1999 or from July 2000 to June 2001, was evaluated. The analysis was restricted to patients with clinically confirmed AMI who underwent testing for both creatine kinase-myocardial band (CK-MB) and troponin. Results were assessed in 3 groups of patients based on biomarker status: those whose findings were positive for troponin only (hereinafter, troponin-only patients), those whose findings were positive for CK-MB only (hereinafter, CK-MB-only patients), and those whose findings were positive for both troponin and CK-MB (hereinafter, troponin/CK-MB patients). Then, the use of guideline-recommended care was compared for patients without contraindications to treatment across the 3 groups. RESULTS: The final study sample included 33 096 patients (mean age, 77.6 years [range, 65-105 years]). The crude in-hospital mortality rate was highest for troponin-only patients (14%) and lowest for CK-MB-only patients (10%, P<.001). After adjusting for demographics, physician specialty, and hospital characteristics, CK-MB-only patients were more likely to receive aspirin (odds ratio [OR], 1.46; 95% confidence interval [CI], 1.28-1.65) and beta-blocker (OR, 1.21; 95% CI, 1.08-1.34) within 24 hours of hospital arrival and aspirin on discharge (OR, 1.27; 95% CI, 1.08-1.49) compared with troponin-only patients. In addition, troponin/CK-MB patients were more likely to receive aspirin (OR, 1.55; 95% CI, 1.42-1.69) and beta-blocker (OR, 1.22; 95% CI, 1.12-1.31) within 24 hours of arrival and on discharge compared with troponin-only patients (ORs, 1.31 [95% CI, 1.17-1.46] and 1.33 [95% CI, 1.15-1.52] for aspirin and beta-blocker, respectively) . CONCLUSIONS: Despite the known poor prognosis associated with troponin elevations in AMI, we demonstrate that guideline-based therapies are underused in older patients with AMI. Therefore, national efforts should focus on the unique characteristics of this high-risk patient population to improve the quality of care for older patients with AMI.

Model-based tracking by classification in a tiny discrete pose space.

A method is presented for tracking 3D objects as they transform rigidly in space within a sparse range image sequence. The method operates in discrete space and exploits the coherence across image frames that results from the relationship between known bounds on the object's velocity and the sensor frame rate. These motion bounds allow the interframe transformation space to be reduced to a reasonable and indeed tiny size, comprising only tens or hundreds of possible states. The tracking problem is in this way cast into a classification framework, effectively trading off localization precision for runtime efficiency and robustness. The method has been implemented and tested extensively on a variety of freeform objects within a sparse range data stream comprising only a few hundred points per image. It has been shown to compare favorably against continuous domain Iterative Closest Point (ICP) tracking methods, performing both more efficiently and more robustly. A hybrid method has also been implemented that executes a small number of ICP iterations following the initial discrete classification phase. This hybrid method is both more efficient than the ICP alone and more robust than either the discrete classification method or the ICP separately.

Validation of bone segmentation and improved 3-D registration using contour coherency in CT data.

A method is presented to validate the segmentation of computed tomography (CT) image sequences, and improve the accuracy and efficiency of the subsequent registration of the three-dimensional surfaces that are reconstructed from the segmented slices. The method compares the shapes of contours extracted from neighborhoods of slices in CT stacks of tibias. The bone is first segmented by an automatic segmentation technique, and the bone contour for each slice is parameterized as a one-dimensional function of normalized arc length versus inscribed angle. These functions are represented as vectors within a K-dimensional space comprising the first K amplitude coefficients of their Fourier Descriptors. The similarity or coherency of neighboring contours is measured by comparing statistical properties of their vector representations within this space. Experimentation has demonstrated this technique to be very effective at identifying low-coherency segmentations. Compared with experienced human operators, in a set of 23 CT stacks (1,633 slices), the method correctly detected 87.5% and 80% of the low-coherency and 97.7% and 95.5% of the high coherency segmentations, respectively from two different automatic segmentation techniques. Removal of the automatically detected low-coherency segmentations also significantly improved the accuracy and time efficiency of the registration of 3-D bone surface models. The registration error was reduced by over 500% (i.e., a factor of 5) and 280%, and the computational performance was improved by 540% and 791% for the two respective segmentation methods.

Validation and improved registration of bone segmentation using contour coherency.

A method is presented to validate the segmentation of computed tomography (CT) image sequences, and im prove the accuracy and efficiency of the subsequent registration of the 3D surfaces that are reconstructed from the segmented slices. The method compares the shapes of contours extracted from neighborhoods of slices in CT stacks of tibias. The bone is first segmented by an automatic segmentation technique, and the bone contour for each slice is parameterized as a 1-D function of normalized arc length versus inscribed angle. These functions are represented as vectors within a K-dimensional space comprising the first K amplitude coefficients of their Fourier Descriptors. The similarity or coherency of neighboring contours is measured by comparing statistical properties of their vector representations within this space. Experimentation has demonstrated this technique to be very effective at automatically identifying low coherency segmentations, the removal of which significantly improved the accuracy and time efficiency of the registration of 3-D bone surface models.