Search

To assess technical and clinical outcomes following lumen-apposing covered self-expanding metallic stent (LAMS) placement for symptomatic pancreatic fluid collections (PFC) with CT and MRI. In this retrospective study, patients with PFC who underwent LAMS placement between March 2015 and June 2018 were included. Primary outcomes included technical success, defined as successful stent placement with resolution of PFC, and clinical success, defined as lack of fluid recurrence after stent removal. Secondary outcomes included time duration from stent placement to removal, complications, and re-intervention need. 28 consecutive patients (20 men, mean age: 53 years ± 17; range 21–75) who underwent endoscopic drainage of symptomatic walled-off necrosis (WON, 21/28, 75%), pseudocyst (PC, 5/28, 18%) or acute necrotic collection (ANC, 2/28, 7%) were included. LAMS were placed successfully in 27/28 (96%) patients. On follow-up imaging after at least one month (n = 24), the volume of the PFC decreased by 99.9% from 425 cm3 [IQR 214 – 636] to 0.6 cm3 [IQR 0–43.9]. After stent removal, 26/27 (96%) patients remained collection free. The median time duration from stent placement to removal was 42 days [IQR 34–71]. Complications (10/28; 36%) included stent occlusion (n = 1), stent migration (n = 3), intraprocedural bleeding (n = 2), postprocedural bleeding (n = 2), and pseudoaneurysm formation (n = 2). Re-intervention was required in 7/27 (26%). Following LAMS placement in patients with symptomatic PFC, high technical and clinical success rates of 96% and 96%, respectively, are achieved. Awareness of common complications seen on cross-sectional imaging might help radiologists and gastroenterologist in the patients’ management.

Clonal hematopoiesis (CH) of indeterminate potential (CHIP), defined as the presence of a somatic mutation in the peripheral blood at a variant allele frequency (VAF) ≥2%, affects at least 10% of individuals older than 65, but low-VAF clones can be detected in 95% of individuals older than 50. CHIP associates with a wide range of comorbidities from atherosclerosis to pulmonary disease. A growing body of evidence, primarily from studies involving Tet2-knockout and stem cell transplant models of CH, suggest that dysregulated inflammation contributes to clonal expansion and associated comorbidities. Mutant leukocytes from animal models contribute to an inflammatory milieu that may confer a selective advantage to the clone, thus perpetuating a cycle of inflammation and expansion. Although it is unclear whether inflammation or expansion sets this cycle in motion, some evidence suggests that inflammation from infections or pre-existing comorbidities initiates this cycle. The pro-inflammatory phenotypes of macrophages from mutant clones and their contributions to disease are well characterized in murine models, but have not yet been confirmed in humans. Furthermore, the roles of other cell types that can carry mutations of CHIP are not fully understood. We propose a rationale for further investigation of neutrophils, other granulocytes and T, B, and NK cells as they may play a role in CHIP-associated comorbidities. As the understanding of CH has advanced, potential interventions, especially those targeting aberrant inflammation, have been proposed. We are hopeful that as studies continue to unravel the complex links between CHIP, inflammation, and leukocyte dysfunction, CHIP-related comorbidities may be more effectively managed.

Corrective interventions while a robot is learning to automate a task provide an intuitive method for a human supervisor to assist the robot and convey information about desired behavior. However, these interventions can impose significant burden on a human supervisor, as each intervention interrupts other work the human is doing, incurs latency with each context switch between supervisor and autonomous control, and requires time to perform. We present LazyDAgger, which extends the interactive imitation learning (IL) algorithm SafeDAgger to reduce context switches between supervisor and autonomous control. We find that LazyDAgger improves the performance and robustness of the learned policy during both learning and execution while limiting burden on the supervisor. Simulation experiments suggest that LazyDAgger can reduce context switches by an average of 60% over SafeDAgger on 3 continuous control tasks while maintaining state-of-the-art policy performance. In physical fabric manipulation experiments with an ABB YuMi robot, LazyDAgger reduces context switches by 60% while achieving a 60% higher success rate than SafeDAgger at execution time., Comment: IEEE CASE 2021

Safety remains a central obstacle preventing widespread use of RL in the real world: learning new tasks in uncertain environments requires extensive exploration, but safety requires limiting exploration. We propose Recovery RL, an algorithm which navigates this tradeoff by (1) leveraging offline data to learn about constraint violating zones before policy learning and (2) separating the goals of improving task performance and constraint satisfaction across two policies: a task policy that only optimizes the task reward and a recovery policy that guides the agent to safety when constraint violation is likely. We evaluate Recovery RL on 6 simulation domains, including two contact-rich manipulation tasks and an image-based navigation task, and an image-based obstacle avoidance task on a physical robot. We compare Recovery RL to 5 prior safe RL methods which jointly optimize for task performance and safety via constrained optimization or reward shaping and find that Recovery RL outperforms the next best prior method across all domains. Results suggest that Recovery RL trades off constraint violations and task successes 2 - 20 times more efficiently in simulation domains and 3 times more efficiently in physical experiments. See https://tinyurl.com/rl-recovery for videos and supplementary material., RA-L and ICRA 2021. First two authors contributed equally

To compare effect of free-text versus structured reporting of IR procedures on report quality and report coding and value.In this retrospective study, 432 common consecutive free-text IR reports created during 4 months (from September 2013 to December 2013) before implementation of structured reporting (February 2014) and 415 structured IR reports created after implementation (from September 2014 to December 2014) were reviewed to assess ease of use and compliance with reporting requirements for regulatory requirements and coding. IR staff and trainees and referring physicians to IR were surveyed on procedure report attributes, such as detail, quality, and clarity.Structured reporting increased compliance with reporting fluoroscopy time, radiation dose, and contrast administration compared with free-text reports (402/432 [93.1%] vs 251/415 [60.5%], P.001; 402/432 [93.1%] vs 242/415 [58.3%], P.001; and 395/432 [91.4%] vs 257/415 [61.9%], P.001). Structured reporting decreased addendum requests for insufficient documentation from 43% (121/435 [28%] to 50/415 [12%], P = .01). Most IR physicians found structured reports to require less time to complete (21/26 [81%]), to be easier to complete (23/26 [89%]), and to have a similar or higher level of detail (19/26 [73%]) compared with free-text reports. Referring physicians were more satisfied with structured reports compared with free-text reports (6.9/10 vs 5.6/10, P = .03).Structured IR reporting compared with free-text reporting improves compliance with radiation dose and contrast reporting, reporting and coding efficiency, and satisfaction among IR and referring physicians.

BACKGROUND: Neutrophils, the most abundant leukocytes and granulocytes, are important regulators of cardiovascular, inflammatory and infectious diseases, yet their role in the pathophysiology of clonal hematopoiesis of indeterminate potential (CHIP) has not been adequately addressed. The effects of inactivating CHIP-driver mutations in the epigenetic regulator TET2 in neutrophils especially, are broadly unknown. HYPOTHESIS: Tet2 inactivation in murine neutrophils, and TET2 mutations in CHIP in humans (CHIP TET2), perturb granulocyte immune effector functions. METHODS: Neutrophils were obtained (EasySep™, StemCell) from the bone marrow of 2- to 4-months-old, sex-matched, control Tet2 f/f;Vav1-icre - (Tet2 f/f) and hematopoietic knockout Tet2 f/f;Vav1-icre + (Tet2 -/-) mice. Neutrophils were cultured (RPMI+10% mouse serum/FBS) and: i) stained with Mitotracker Deep Red/Nuc Blue, co-cultured and imaged (Leica SP8-X) for 30min with GFP-labeled Staphylococcus aureus (10:1 ratio) and analyzed in FIJI; ii) cultured for 3h with vehicle or 10μg/mL of S. aureus lipotechoic acid (LTA). RNA-Seq was generated (Illumina QuantSeq 3' mRNA, single-end 75bp read lengths, 5 million reads/sample), trimmed, aligned to GRCm39 using STAR. CHIP participant DNA and RNA were sequenced previously from whole blood (Cook et al., Bld Adv 2019; Cook et al., ASH 2018, with a 48-gene panel on Ion Proton, and ribo-depleted bulk RNA on Illumina, respectively). New CHIP TET2 vs. no CHIP, and murine RNA-Seq analyses were carried out in DESeq2. Human serum granule protein levels were quantified by ELISA (VersaMax). Mann-Whitney U tests were carried out in Prism. P RESULTS: Tet2 -/- mice had 1.34-fold more bone marrow CD11b +Ly6G + neutrophils than control Tet2 f/f mice (p=0.03), consistent with myeloid expansion. Compared to Tet2 f/f, Tet2 -/- neutrophils phagocytosed fewer S. aureus (Fig1A) and moved more slowly (Fig1B). Preliminary data suggest that Tet2 -/- neutrophil extracellular trap (NET) formation in response to S. aureus was also impaired, showing fewer and less extensive NETs (Fig1C). LTA-stimulated gene expression profiles were similar between Tet2 -/- and Tet2 f/f, suggesting pre-existing differences at baseline. Unexpectedly, the most significant GO term enrichment related to upregulated viral response pathways, including interferon-stimulated genes, (e.g. Ifitm1). The cause is unknown, but this is reminiscent of the constitutive interferon response seen in myelodysplastic syndrome (MDS) patients and TET2-mutant hematopoietic stem cells, where epigenetic dysregulation of endogenous retrotransposable elements leads to a viral mimicry response. Tet2 -/- neutrophils also overexpressed Asprv1, a regulator of inflammation ostensibly acquired from a retrotransposon. Interestingly, Ccdc80, which has been linked to Tet2 and Jak2 functions, was most significantly downregulated in Tet2 -/-, along with the Pnpla1 lipid phosphatase. Finally, Tesc, a promoter of granulocytic differentiation, was upregulated in Tet2 -/-, and there were perturbations of genes encoding neutrophil granule contents. Similarly, human RNA-Seq revealed that several leukocyte (de)granulation-related genes (e.g. lactoferrin LTF, myeloperoxidase MPO) were upregulated in CHIP TET2 subjects to those without CHIP, and these corresponded with higher LTF and MPO serum titers in an expanded cohort (Fig1D,E). Finally, there were striking decreases of gene expression associated with cytotoxic (T/NK) human lymphocytes (i.e. GZMM, TRGV8, etc.). Neutrophil, lymphocyte and monocyte counts were not significantly different between the groups. CONCLUSIONS: Tet2-deficient murine neutrophils have compromised immune function, possibly due to differences in pre-stimulus state. TET2-mutation carrying neutrophils in CHIP may exhibit similar abnormalities, as has been previously noted in neutrophils isolated from MDS patients. Indeed, CHIP is now known to associate with increased risk of bacterial and viral infections, and infection risk has also previously been noted for MDS. People with CHIP have elevated peripheral blood serum MPO and LTF levels, suggesting a difference in leukocyte granule biology, likely related to neutrophils. These data aid in understanding how CHIP alters immunity. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

This article details findings from a multidisciplinary research team’s inquiry into the social uses of 3D printing. It applies digital research methods to 3D printing communities and their digitally shared objects. Thingiverse is one of the most well-known file repositories available for the semi-public distribution of files for use in 3D printers. The site allows for several different means of metadata classification of these files. Previous research on the site focused on the legal concerns related to the different types of intellectual property license ‘metadata’ attached to objects. Beyond these data points are numerous additional types of elected and automated connections between both users, objects and processes of creation including liking, commenting, tagging, categorising, watching, collecting, remixing, making, sharing, and attribution. Social Network Analysis of the relations between these data reveal interesting patterns towards the use and life of these objects. Our analysis also show...

PURPOSE To describe the implementation and efficacy of electronic capture of interventional radiology (IR) group peer review performed during morning rounds in an academic institution. MATERIALS AND METHODS Peer-review submissions 24 months before (July 2012 through June 2014) and 24 months after (July 2014 through June 2016) implementation of an electronic group IR peer review were evaluated. Electronic IR peer review assessed clinical decision-making, patient care, and technique appropriateness by using the following responses: agree, acceptable alternative, or inappropriate. Complications, near-misses, and learning opportunities were also noted, and explanations were provided in the comments. An IR attending physician documented the consensus discussion via an online form or mobile app. RESULTS After implementation of electronic capture of IR group peer reviews, 9.5% of all IR procedures (773 of 8,152) were reviewed, compared with 0.9% (63 of 7,152) before the implementation (P < .0001). On average, 3.4 ± 0.5 IR attending physicians participated in each review, and 17.3 ± 9.7 IR cases were reviewed per month by each attending physician. Peers fully agreed with the chosen technique in 729 of 773 cases (94.3%) and noted an acceptable alternative technique in 40 of 773 (5.2%). In 4 cases (0.5%), the chosen technique was rated as inappropriate. Peers fully agreed with chosen clinical decision-making and patient care in 765 of 773 cases (99%), with 8 cases (1%) deemed as acceptable alternatives. Learning opportunities were documented in 9.6% of cases (74 of 773). CONCLUSIONS Electronic IR-specific group peer review captures the clinical decision-making process, patient outcomes, technical appropriateness, and short-term complications and may provide practice-improvement solutions.

The purpose of this study was to evaluate clinical, functional, and radiographic outcomes following total knee arthroplasty (TKA) performed with patient-specific instrumentation (PSI), computer-assisted surgery (CAS), and manual instruments at short-term follow-up.122 TKAs were performed by a single surgeon: 42 with PSI, 38 with CAS, and 40 with manual instrumentation. Preoperative, 1-month, and 6-month clinical and functional outcomes were measured using the Knee Society scoring system (knee score, function score, range of motion, and pain score). Improvements in clinical and functional outcomes from the preoperative to postoperative period were analyzed. Preoperative and postoperative radiographs were measured to evaluate limb and component alignment.Preoperative, 1-month postoperative, and 6-month postoperative knee scores, function scores, range of motion, and pain scores were highest in the PSI group compared to CAS and manual instrumentation. At 6-month follow-up, PSI TKA was associated with a statistically significant improvement in functional score when compared to manual TKA. Otherwise, there were no statistically significant differences in improvements among PSI, CAS, and manual TKA groups.The higher preoperative scores in the PSI group limits the ability to draw definitive conclusions from the raw postoperative scores, but analyzing the changes in scores revealed that PSI was associated with a statistically significant improvement in Knee Society Functional score at 6-month post-TKA as compared to CAS or manual TKA. This may be attributable to improvements in component rotation and positioning, improved component size accuracy, or other factors that are not discernible on plain radiograph.

The purpose of this study was to determine whether differences in clinical, functional, or radiographic outcomes existed at 5-year follow-up between patients who underwent computer-assisted or manual total knee arthroplasty (TKA). Seventy-eight consecutive TKAs were performed by a single surgeon who had extensive experience performing computer-assisted and manual TKA. The manual group (n=40) and computer-assisted group (n=38) were similar with regard to age, sex, diagnosis, body mass index, surgical technique, implants, perioperative management, Knee Society scores, and anteroposterior mechanical axis. Sixty-three (manual group, n=34; computer-assisted group, n=29) patients were available for final follow-up. At 5-year follow-up, no statistically significant differences were found in Knee Society knee score ( P =.289), function score ( P =.272), range of motion ( P =.284), pain score ( P =.432), or UCLA activity score ( P =.109) between the 2 groups. Postoperative radiographs showed a significant difference in the mechanical axis ( P =.004) between the 2 groups; however, both groups achieved a neutral mechanical axis of ±3° (computer-assisted group mean, 2.0°; manual group mean, −0.24°). When TKA was performed by an experienced surgeon, no significant difference was identified at 5-year follow-up between patients who underwent computer-assisted vs manual TKA.