Uncertain Cases Research Articles

A Novel Non-Probabilistic Sensor Placement Method for Structural Health Monitoring Using an Iterative Multiobjective Optimization Algorithm

A novel non-probabilistic sensor placement method for structural health monitoring is proposed with interval numbers based on the relationship of interval for sensor number (RISEN) index, and an iterative multiobjective optimization algorithm is constituted to optimize sensor placement. To avoid the limitation of scarce statistical information, the interval objectives of effective independence and eigenvalue vector product methods are derived. To overcome the inaccuracy of sensor number decisions using the determinate methods in uncertain cases, the novel RISEN index is defined to ascertain the best sensor number. Considering the number and locations of sensors as two types of design variables, two methods are regarded as optimization objectives, which are composed of the multiobjective optimal sensor placement methods. Based on the modified hypervolume evaluation index, an iterative multiobjective optimization algorithm is investigated using the updating process to improve the efficiency of the sensor placement. The validity of the method is proven using four examples.

1029 FABRY DISEASE: A RARE CAUSE TO REMEMBER WHEN FACING ACUTE KIDNEY INJURY IN YOUNG

Abstract Introduction Anderson-Fabry disease (FD) is a rare X-linked hereditary disease caused by mutations in the alpha-galactosidase A (GLA) gene, a lysosomal hydrolase that catabolizes lipids. GLA deficency leads to a progressive accumulation of undegraded glycosphingolipids, mainly Globotriaosylceramide (GB3), within lysosomes of cells (epithelial and endothelial cells, neurons, cardiomyocytes and renal cells), leading to cellular dysfunction. The incidence of FD is 1:117.000 but in patients with end stage renal disease (ESRD), incidence is higher, ranging from 0.04% up to 1.16% in male dialysis patients. Classical form of the disease shows renal, cardiac and cerebrovascular involvements usually. Cardiac involvement includes hypertrophic cardiomyopathy, arrhythmias such as complete heart block, valve dysfunction and myocardial infarction. Renal manifestations include proteinuria, progressive loss of renal function with chronic kidney disease (CKD) that leads to ESRD. The diagnosis of FD can be very difficult and often is delayed due to subtle clinical manifestations, but an early diagnosis is crucial to start enzyme replacement therapy (ERT), based on recombinant human GLA, as soon as possible. Clinical Case we describe the case of a 36-years-old male admitted to the Emergency Room of our hospital due to asthenia, nausea, dysphagia, anuria and loss of weight. At physical examination, apical systolic murmur and a bilateral ankle edema. Blood pressure was 195/100 mmHg, heart rate 125 beats per minute, peripheral oxygen saturation 100%. Blood test showed Hemoglobin 8 g/dL, Blood Urea 422 mg/dL, Creatinine 20 mg/dL, Na+ 135 mEq/L, K+ 4 mEq/L, Phosphorus 9 mg/dL. Arterial blood gas analysis showed metabolic and lactic acidosis. The diagnosis of acute kidney injury (AKI) was made, and a dialytic treatment was scheduled. During dialytic session, the patient had a syncope. EKG showed a complete left bundle branch block (LBBB) with advanced atrio-ventricular block (2:1, complete). An echocardiogram highlighted a severe myocardial hypertrophy (septal wall 25 mm) involving also the right ventricle with “ground glass” aspect of the posterolateral, posterior and inferior septum, with minimal pericardial effusion. Heart MRI scan confirmed LV myocardial hypertrophy with circumferential aspect (interventricular septum 26.6 mm), mild decrease in left ventricular ejection fraction (47%) and myocardial hypointensity area in the FSE sequences. Renal ultrasound showed bilateral cortical hyperecogenity, small cortico-medullary border, initial reduction of kidney volume and diffuse perirenal edema. In consideration of the high probability of a storage disease, we performed genetic testing for FD and Gaucher disease, peri-umbilical fat biopsy for Amyloid Disease and a renal biopsy. The genetic test resulted positive for typical mutation of FD and renal biopsy documented ESRD secondary to glycosphingolipid storage disease ceramid type from FD. Conclusion Differential diagnosis of AKI in a young male should include FD, as suggested by the clinical case reported, and nephrologist should screen young patients for α-GLA enzyme deficiency (diagnostic for FD) possibly coupled with renal biopsy in uncertain cases.

2018. Comparison of Trends in Hospital-Onset Bloodstream Infections (HOBSIs) and Central Line Associated Bloodstream Infections (CLABSIs) across a Three-Hospital Health System in the COVID Era

Abstract Background Hospital-onset bloodstream infection (HOBSI) incidence has been proposed as a complementary quality metric to central line-associated bloodstream infection (CLABSI) surveillance. Several recent studies have detailed increases in median HOBSI and CLABSI rates during the COVID-19 pandemic. We sought to understand trends in HOBSI and CLABSI rates at a single health system in the context of COVID-19. Methods We conducted a retrospective analysis of HOBSIs and CLABSIs at a three-hospital health system from 2017 to 2021 (Figure 1). We compared counts, denominators, and demographic data for HOBSIs and CLABSIs between the pre-pandemic (1/1/2017-3/30/2020) and pandemic period (4/1/2020-12/31/2021) (Table 1). We applied Poisson or negative binomial regression models to estimate the monthly change in incidence of HO-BSI and CLABSI rates over the study period. Figure 1:Definitions applied for hospital-onset bloodstream infections (HO-BSIs) and central line-associated bloodstream infections (CLABSIs). Potentially contaminated blood cultures were identified by microbiology laboratory technicians as any set of blood culture in which a single bottle was positive for organisms typically considered as skin contaminants. Uncertain cases undergo secondary review by senior lab technicians. Table 1:Count, denominator, and device utilization ratio data for hospital-onset bloodstream infections (HO-BSIs) and central line-associated bloodstream infections (CLABSIs)Note that central line utilization increased upon regression analysis (p<0.001). Results The median monthly HOBSI rate per 1,000 patient days increased from 1.0 in the pre-pandemic to 1.3 (p< 0.01) in the pandemic period, whereas the median monthly CLABSI rate per central line days was stable (1.01 to 0.88; p=0.1; Table 2). Our regression analysis found that monthly rates of HO-BSIs increased throughout the study, but the increase was not associated with the onset of the COVID-19 pandemic based on comparisons of model fit (Figure 2; Table 3). Despite an increase in central line utilization, regression modelling found no changes in monthly CLABSIs rates with respect to time and the COVID-19 pandemic. Incidence of HOBSIs and CLABSIs by common nosocomial organisms generally increased over this time period, though time to infection onset remained unchanged in our studied population (Table 2). Table 2:Epidemiology and demographic characteristics of hospital-onset bloodstream infections (HO-BSIs) and central line-associated bloodstream infections (CLABSIs).Figure 2:Regression analysis of monthly rates of HO-BSIs and CLABSIs over time per patient days (left) and CLABSIs per central line days (right). Gray areas denote COVID-19 pandemic period (April 2020 to December 2021). Provided p values refer to the statistical significance of the coefficient of the time variable in each model. Table 3:Likelihood ratio tests comparing regression models.No statistical significance was noted between models that included COVID-19 variables (including a time-trended post-pandemic variable) and models that were only time-dependent. Conclusion HOBSIs rates did not correlate with CLABSI incidence across a three-hospital health system from 2017 and 2021, as rates of HOBSI increased but CLABSI rates remained flat. Our observed increase in HOBSI rates did not correlate with the onset of the COVID-19 pandemic, and caution should be used in modeling the effects of COVID-19 without time-trended analysis. Further evaluation is needed to understand the etiology, epidemiology, and preventability of HO-BSI. Disclosures All Authors: No reported disclosures.

Ambiguous melanocytic lesions: A retrospective cohort study of incidence and outcome of melanocytic tumor of uncertain malignant potential (MELTUMP) and superficial atypical melanocytic proliferation of uncertain significance (SAMPUS) in the Netherlands

Melanocytic tumor of uncertain malignant potential (MELTUMP) and superficial atypical melanocytic proliferation of uncertain significance (SAMPUS) are descriptive and provisional terms for melanocytic tumors with ambiguous histopathological features that are not easily classified as either benign or malignant. To investigate the incidence and clinical outcome of MELTUMP and SAMPUS in the Netherlands. In this retrospective cohort study, we reviewed all diagnoses of MELTUMP and SAMPUS from the Dutch Nationwide Pathology Databank from 1991 to October 1, 2021. Clinical outcome was studied for cases diagnosed until October 1, 2018. A total of 1685 MELTUMP and 1957 SAMPUS were identified with an annual incidence of 150 to 300 cases. Metastatic behavior was seen in 0.7% of all initially diagnosed MELTUMP. All SAMPUS remained free of metastases. Reassessment of pathology slides and confirmation of clonality between primary and metastatic lesions remained outside the scope of this study. Despite the 'uncertainty' in the nomenclature, our results demonstrate a low malignant potential for MELTUMP and no malignant potential for SAMPUS. We emphasize the importance of consultation for ambiguous melanocytic lesions and to limit the MELTUMP/SAMPUS terminology to legitimately uncertain or unclassifiable cases.

Giant Pilomatricoma Mistaken for a Malignant Tumor.

Pilomatricomas are tumors originating from the matrices of hair follicles. Giant pilomatricomas, defined as pilomatricomas that are 5cm or larger, are benign but may appear malignant clinically. We present the case of a 69-year-old man with a rapidly growing mass on his scalp. When he visited our department, the tumor measured 10.0×6.0×4.0cm and showed inflammation and ulceration. Magnetic resonance imaging and 18 F-fluorodeoxyglucose positron emission tomography-computed tomography showed findings resembling lymph node metastasis from a malignant tumor. However, upon an incisional biopsy, the tumor was diagnosed as a pilomatricoma. Therefore, we performed an excisional biopsy instead of radical surgery and lymph node dissection. The tumor was ultimately diagnosed as a giant pilomatricoma based on the excisional biopsy, and the patient received reconstruction only at the site of the defect. A giant pilomatricoma can be mistaken for a malignant tumor due to its characteristics. In such uncertain cases, it can be helpful to first perform an excisional biopsy.

Using Radiomics to Improve the Diagnostic Accuracy of Indeterminate Residual Primary Disease on Restaging PET/CT Imaging Following Radiation Therapy for Head and Neck Cancers

<h3>Purpose/Objective(s)</h3> Identifying residual disease (RD) in the primary site of head and neck cancer (HNC) patients treated with radiotherapy (RT) on the 3-month post RT PET/CT scans can be challenging. Radiologists have to account for RT-related inflammation versus RD. Accurate identification is key to determine the need for morbid salvage versus close observation. In this study, we hypothesized a clinical/radiomics model could complement a radiologist's interpretation in these uncertain cases. <h3>Materials/Methods</h3> We identified 182 patients with HNC treated at our institution from 2016-2020. We reviewed the 3-month post-RT PET/CT scan interpretations to obtain the radiologists' RD assessment. We classified assessments as "certain" or "uncertain." "Certain" assessments included words like "complete metabolic response," "no RD," "RD present" while "uncertain" assessments provided a differential. For uncertain cases, predictions were inferred as the most likely diagnosis ("likely radiation changes but can't rule out RD" was no RD present). If no clarification was provided, then the prediction was marked as RD present. We extracted radiomics features from the CT and PET modalities using open-source software. We gathered clinical data such as HPV status, T stage, etc. Feature reduction was performed by assessing single feature performance on a logistic regression model. Features with high correlation (>0.9) measured using Pearson coefficient were discarded. Recursive feature elimination was performed on the remaining features (only 5-10 features were allowed in the final model). We trained an explainable boosting machine model using the "certain" data and tested with "uncertain" data. Ground truth was obtained from follow-up imaging or pathology. <h3>Results</h3> Ninety-six assessments were "certain" and 86 were "uncertain." RD was present in 10 and 9 cases, respectively. The performances for the radiologists and our models are summarized in Table 1. The Clinical/CT radiomics model reduced the number of false positives from 22 to 8 at the cost of only one additional false negative. The ensemble of our CT and PET models did not improve performance. Table 1 <h3>Conclusion</h3> Even expert radiologists may struggle with identifying RD given post-RT inflammatory changes. Our Clinical/CT radiomics model greatly improved specificity with only a minor drop in sensitivity in this uncertain cohort. Machine learning models may have utility in complementing clinical experience to improve diagnostic accuracy for this challenging scenario.

Multiple Mobile Robots Coordination in Shared Workspace for Task Makespan Minimization

In this paper we consider a system of multiple mobile robots (MMRS) and the process of their concurrent motion in a shared two-dimensional workspace. The goal is to plan the robot movement along given fixed paths so as to minimize the completion time of all the robots while ensuring that they never collide. Thus, the considered problem combines the problems of robot schedule optimization with collision and deadlock avoidance. The problem formulation is presented and its equivalent reformulation that does not depend explicitly on the geometry of the robot paths is proposed. An event-based solution representation is proposed, allowing for a discrete optimization approach. Two types of possible deadlocks are identified and deadlock avoidance procedures are discussed. We proposed two types of solving methods. First, we implemented two metaheuristics: the local-search-based taboo search as well as the population-based artificial bee colony. Next, we implemented 14 simple constructive algorithms, employing dispatch rules such as first-in first-out, shortest distance remaining first, and longest distance remaining first, among others. A set of problem instances for different numbers of robots is created and provided as a benchmark. The effectiveness of the solving methods is then evaluated by simulation using the generated instances. Both deterministic and lognormal-distributed uncertain robot travel times are considered. The results prove that the taboo search metaheuristic obtained the best results for both deterministic and uncertain cases, with only artificial bee colony and a few constructive algorithms managing to remain competitive. Detailed results as well as ideas to further improve proposed methods are discussed.

Neck-shaft angle measurement in children: accuracy of the conventional radiography-based (2D) methods compared to 3D reconstructions

Aim of this study was to examine the accuracy of widely used conventional radiography-based (2D) neck-shaft angle measurements compared to 3D reconstruction. In our retrospective study, EOS 2D/3D images of 156 patients (312 limbs) were selected from our database (4–16 years old: 6 girls and 6 boys/year), where no pathology was revealed. Using the 2D modality of the EOS method neck-shaft angle was measured using the “biggest diameter” and “circle fitting” techniques to define the femoral neck axis and 1/3, 1/2 and full femur to determine the femoral shaft axis. EOS 3D reconstructions of same images were also performed and a comparison of 2D and 3D results was made. We did not find any significant difference between accuracy of the four examined 2D methods, although the deviation between 2 and 3D results was considerable (average difference: 5.11–5.58°, p < 0,001). In 31% of the cases, difference was more than 10°. Only femoral torsion showed significant influence on the difference (correlation coefficient: 0.380, p < 0.001). We did not find a clinically significant difference between the examined 2D methods, although their accuracy was highly questionable compared to 3D results. We suggest using any 3D imaging method for surgical planning and in uncertain cases.

Characterizing and learning from medical malpractice cases against medical oncologists using a national claims database.

318 Background: Malpractice cases may contain valuable information to inform safety improvement efforts in cancer care. No analyses of malpractice cases have been published focusing on medical oncologists as the primary defendant. The recent development of a new safety incident taxonomy for medical oncology enabled us to classify malpractice cases and identify patient safety improvements.1. Methods: We conducted a qualitative analysis of case vignettes naming medical oncologists as the primary defendant. Case vignettes were obtained from CRICO’s Candello database. Of 452 cases against medical oncologists, a random sample of 100 cases was selected. Cases were coded for the type of safety incident, and whether a system solution could plausibly reduce the risk of a future event. Coding was performed by 3 physicians with experience caring for patients with cancer. Two reviewers reviewed each case; discrepancies were resolved by the third reviewer. Results: Ninety-nine cases were included in the final analysis; 1 was determined to be misassigned to medical oncology as the primary defendant when the case vignette was entirely about radiation oncology treatment and side effects. The most frequent cancer types were breast (16%), acute leukemia (9%), aggressive lymphoma (8%), and gynecologic malignancies (7%). The most common safety incidents identified were provider clinical management errors (41%), adverse drug reactions (20%), relational issues among providers, patients, and other healthcare staff (13%), and prescriber ordering errors (6%). 47 case vignettes described safety incidents that might have been prevented by a systems intervention. The most common interventions were “tumor board or expert review of rare or uncertain cases” (19%), “pharmacist review for safe prescribing of chemotherapy” (17%), “cancer navigator for appointment scheduling in high-risk settings” (9%), “closed loop communication with radiology” (6%), and “electronic patient portal access to all test results” (6%). Conclusions: Using a new medical oncology-focused incident coding taxonomy, we were able to classify all 99 malpractice cases. Human factors issues accounted for the majority of claims. We determined that more comprehensive use of existing improvement interventions might substantially reduce the risk of future events. 1 Jacobson, J. et al. Development of a taxonomy for characterizing medical oncology-related patient safety and quality incidents: A novel approach. BMJ Open Quality. [Manuscript in press]. Accepted June 10, 2022. DOI 10.1136/bmjoq-2022-001828.R2.

Rapid evaluation of T cell clonality in the diagnostic work-up of mature T cell neoplasms: TRBC1-based flow cytometric assay experience

The identification of mature T cell neoplasms by flow cytometry is often challenging, due to overlapping features with reactive T cells and limitations of currently available T cell clonality assays. The description of an antibody specific for one of two mutually exclusive T cell receptor (TCR) β-chain constant regions (TRBC1) provides an opportunity to facilitate the detection of clonal TCRαβ+ T cells based on TRBC-restriction. Here we prospectively analyzed 14 healthy controls and 63 patients with the flow cytometry protocol currently used for suspected T cell neoplasm implemented with immunostaining targeting TRBC1. Specimens were firstly classified in 3 groups based on clinical records data, laboratory findings and immunophenotypic features. T cell clonality was assessed by TCR Vβ repertoire analysis and the new rapid TRBC1 assay. Results showed that TRBC1 unimodal expression was unequivocally associated with samples presenting with immunophenotypic aberrancies. Moreover, we demonstrated that the use of TRBC1 is useful in solving uncertain cases and confirmed the high sensitivity of the method in identifying small T cell clones of uncertain significance (T-CUS). Finally, we found a high degree of concordance (97%) comparing the currently available clonality assessment methods with the proposed new method. In conclusion, our results provided real-life evidence of the utility of TRBC1 introduction in the flow cytometric clonality evaluation for the routine diagnostic work-up of T cell neoplasms.

Secondary attack rate of the COVID-19

Introduction: The first case of COVID-19 infection in Kazakhstan was registered on March 13, 2020. At the beginning the detection methods and the load of the spread of the emerging respiratory pathogen were uncertain. This study aimed to assess the incidence of the secondary attack rate among close contacts of confirmed and probable COVID-19 cases living in the same household in Nur-Sultan, Kazakhstan. Methods: The prospective study included 172 participants: 122 confirmed and 50 uncertain cases of COVID-19 with varying degrees of severity as well as their close contacts identified in Nur-Sultan, Kazakhstan from November 26th, 2020 until February 15th, 2021. All participants were tested with PCR and ELISA assays at the time of inclusion and on days 14-21 of follow-up. Results: The most common symptoms among both cohorts of patients were fever (90%) (p=&lt;0.001), cough (78.0%) (p=0.11), difficulty breathing (63.3%) (p=&lt;0.001), chills (60%) (p=&lt;0.001). The effective reproductive Rt number for confirmed cases was 1.43 [95%CI=1.27-1.59], for probable cases 0.96 [95%CI=0.70-1.22]. The mean serial interval was 4.02 (SD 2.52), for confirmed was 4.43 (SD 2.45), for probable cases was 3.21 (SD 2.48). Conclusion: The infection rates following close contact with COVID-19 confirmed and probable cases were 92% and 98.1%, respectively.

DNAJB9 Is a Reliable Immunohistochemical Marker of Fibrillary Glomerulonephritis: Evaluation of Diagnostic Efficacy in a Large Series of Kidney Biopsies

Fibrillary glomerulonephritis (FGN) is a rare glomerular disease characterized by a challenging diagnostic workup requiring ultrastructural identification of 20 nm-thick randomly oriented fibrillar deposits. However, the recent introduction of DNAJB9 as a putative diagnostic marker of FGN could thoroughly improve this diagnostic scenario. This study aims to assess the DNAJB9 immunohistochemical expression in a large series of FGN cases and to eventually confirm its role as a diagnostic marker of FGN. We evaluated the immunohistochemical expression of DNAJB9 (Rabbit Polyclonal, ThermoFisher) in a series of 77 FGN and 128 non-FGN cases diagnosed between January 1992 and June 2022 at the Pathology Unit of the AOU Città della Salute e della Scienza Hospital. DNAJB9 was expressed in 73 of the 74 evaluable FGN cases, mostly showing a strong glomerular positivity (68 cases). Additionally, DNAJB9 resulted positive in all challenging scenarios [early-stage (6), congophilic (4), combined (4), and uncertain (4) cases of FGN)]. DNAJB9 was negative in all non-FGN cases, eventually resulting in a specificity of 100% and sensitivity of 99%. In conclusion, we confirmed the role of DNAJB9 as a diagnostic marker of FGN. Its adoption in the clinical routine will allow a faster, more feasible, and more accurate FGN diagnosis.

Detection and Classification of Bronchiectasis Based on Improved Mask-RCNN.

Bronchiectasis is defined as a permanent dilation of the bronchi that can cause pulmonary ventilation dysfunction. CT examination is an important means of diagnosing bronchiectasis. It can also be used in severity scoring. Current studies on bronchiectasis have focused on high-resolution CT (HRCT), ignoring the more common low-dose CT (LDCT). Methodologically, existing studies have not adopted an authoritative standard to classify the severity of bronchiectasis. In effect, the accuracy of detection and classification needs to be improved for practical application. In this paper, the ACER image enhancement method, RDU-Net lung lobe segmentation method and HDC Mask R-CNN model were proposed to detect and classify bronchiectasis. Moreover, a Python-based system was developed: after inputing an LDCT image of a patient’s lung, it can automatically perform a series of processing, then call on the trained deep learning model for detection and classification, and automatically obtain the patient’s bronchiectasis final score according to the Reiff and BRICS scoring criteria. In this paper, the mapping relationship between original lung CT image data and bronchiectasis scoring system was established. The accuracy of the method proposed in this paper was 91.4%; the IOU, sensitivity and specificity were 88.8%, 88.6% and 85.4%, respectively; and the recognition speed of one picture was about 1 s. Compared to a human doctor, the system can process large amounts of data simultaneously, quickly and efficiently, with the same judgment accuracy as a human doctor. Doctors only need to judge the uncertain cases, which significantly reduces the burden of doctors and provides a useful reference for doctors to diagnose the disease.

Stochastic dynamic programming based optimal energy scheduling for a hybrid fuel cell/PV/battery system under uncertainty

Hybrid green energy system, consisting of photovoltaic (PV), fuel cell and battery, receives wide attention because of its autonomy, flexibility and promising potential in accelerating the development of carbon neutrality in the field of power generation. However, the efficient power dispatching of the hybrid energy system is challenging due to the inevitable uncertainties of the solar energy. To this end, stochastic dynamic programming (SDP) is used in this paper to find the optimal solution to minimize the total fuel consumption during a 72-hour operating cycle, with constraints on the magnitude and rate of the fuel cell and battery operation, in which the stochastic characteristics of the solar power are described by Markov chain. The influence of sampling time and number of states on the solar prediction accuracy is discussed. For comparison, the traditional rule-based algorithm and dynamic programming (DP) algorithms are also utilized to describe and solve the power distribution problem, corresponding to different decision results in terms of how to distribute the energy flow for each time period. Numerical optimization results within the 72-hour period demonstrate that, the SDP has a 20.61% economy and 66.34% battery SOC improvement than that of rule-based algorithm, and in most uncertain cases, the SDP produces superior economic performance than those of both the rule-based algorithm and DP, benefited from the inclusion of the solar power probabilities into the optimization framework. The results of this paper lay a solid foundation for the efficient energy management of the hydrogen and solar hybrid energy system.

Diagnostic Accuracy of FDG PET for the Identification of Vascular Graft Infection

Fluorodeoxyglucose (FDG) positron emission tomography (FDG PET/CT) can be used to identify and localize infection in patients with vascular graft infections (VGI). We aimed to evaluate the diagnostic accuracy of 18F-FDG PET/CT by defining thresholds for standardized uptake value (SUV) and tissue-to-background ratio (TBR) that would accurately identify the presence of vascular graft infection. Patients with suspected VGI were prospectively recruited and underwent 18F-FDG PET/CT scans. Diagnosis was based on clinical, laboratory and radiologic findings, and blinded to the results of the PET/CT scan. Receiver operator characteristics (ROC) curve analysis was done to determine optimal thresholds for SUV and TBR. Our final cohort consisted of 28 patients with suspected VGI (mean±SD age 67±10years, 61% men), of which 15 patients (54%) had definitive VGI. The cohort included 61% prosthetics grafts and 39% stent-grafts. The type of graft included in this study were biologic (4%), Dacron (64%) and Polytetrafluoroethylene (32%). The location of the implanted grafts was aortic (54%) and peripheral arterial reconstruction (46%). The location of the peripheral graft was 77% in lower extremity and 23% in the upper extremity (arterio-venous grafts for dialysis access). Using ROC analysis, SUV max of 4.5, SUV mean of 3.7, and a TBR of 1.6 gave the best balance between sensitivity and specificity (93%/92%, 100%/92% and 93%/92%, respectively). All thresholds had an area under the curve ≥0.93 and correct reclassification rate ≥93%. Our data suggests that FDG PET/CT can be used to reliably and accurately diagnose VGI. The dual anatomic-physiologic information from FDG PET/CT can complement clinical diagnosis particularly in uncertain cases.

Deep Learning Fuzzy Inference: An Interpretable Model for Detecting Indirect Immunofluorescence Patterns Associated with Nasopharyngeal Cancer

The detection of serum Epstein-Barr virus antibodies by immunofluorescence assay (IFA) is considered the gold standard screening test for nasopharyngeal cancer (NPC) in high-risk populations. Given the high survival rate after early detection in asymptomatic patients, compared to the poor prognosis in patients with late-stage NPC, screening using IFA has tremendous potential for saving lives in the general population. However, IFA requires visual interpretation of cellular staining patterns by trained pathology staff, making it labor intensive and hence nonscalable. In this study, an automated fuzzy inference (FI) system achieved high agreement with a human IFA expert in identifying cellular patterns associated with NPC (κ=0.82). The integration of a deep learning module into FI further improved the performance of FI (κ=0.90) and reduced the number of uncertain cases that required manual evaluation. The performance of the resulting hybrid model, termed deep learning FI (DeLFI), was then evaluated with a separate testing set of clinical samples. In this clinical validation, DeLFI outperformed human evaluation on the area under the curve (0.926 versus 0.821) and closely matched human performance on Youden J index (0.81 versus 0.80). Data from this study indicate that the combination of deep learning with FI in DeLFI has the potential to improve the scalability and accuracy of NPC detection.

Secreted miR-153 Controls Proliferation and Invasion of Higher Gleason Score Prostate Cancer

Prostate cancer (PC) is a male common neoplasm and is the second leading cause of cancer death in American men. PC is traditionally diagnosed by the evaluation of prostate secreted antigen (PSA) in the blood. Due to the high levels of false positives, digital rectal examination and transrectal ultrasound guided biopsy are necessary in uncertain cases with elevated PSA levels. Nevertheless, the high mortality rate suggests that new PC biomarkers are urgently needed to help clinical diagnosis. In a previous study, we have identified a network of genes, altered in high Gleason Score (GS) PC (GS ≥ 7), being regulated by miR-153. Until now, no publication has explained the mechanism of action of miR-153 in PC. By in vitro studies, we found that the overexpression of miR-153 in high GS cell lines is required to control cell proliferation, migration and invasion rates, targeting Kruppel-like factor 5 (KLF5). Moreover, miR-153 could be secreted by exosomes and microvesicles in the microenvironment and, once entered into the surrounding tissue, could influence cellular growth. Being upregulated in high GS human PC, miR-153 could be proposed as a circulating biomarker for PC diagnosis.

Uncertainty Quantification in Segmenting Tuberculosis-Consistent Findings in Frontal Chest X-rays.

Deep learning (DL) methods have demonstrated superior performance in medical image segmentation tasks. However, selecting a loss function that conforms to the data characteristics is critical for optimal performance. Further, the direct use of traditional DL models does not provide a measure of uncertainty in predictions. Even high-quality automated predictions for medical diagnostic applications demand uncertainty quantification to gain user trust. In this study, we aim to investigate the benefits of (i) selecting an appropriate loss function and (ii) quantifying uncertainty in predictions using a VGG16-based-U-Net model with the Monto–Carlo (MCD) Dropout method for segmenting Tuberculosis (TB)-consistent findings in frontal chest X-rays (CXRs). We determine an optimal uncertainty threshold based on several uncertainty-related metrics. This threshold is used to select and refer highly uncertain cases to an expert. Experimental results demonstrate that (i) the model trained with a modified Focal Tversky loss function delivered superior segmentation performance (mean average precision (mAP): 0.5710, 95% confidence interval (CI): (0.4021,0.7399)), (ii) the model with 30 MC forward passes during inference further improved and stabilized performance (mAP: 0.5721, 95% CI: (0.4032,0.7410), and (iii) an uncertainty threshold of 0.7 is observed to be optimal to refer highly uncertain cases.

Dismantling antibiotic infrastructures in residential aged care: The invisible work of antimicrobial stewardship (AMS)

Recent social science scholarship has sought to understand the visible and invisible impacts of how antibiotics are entrenched as infrastructures and put to work as a proxy for higher levels of care (clinical or otherwise) within modern healthcare. Using a qualitative research design, in this paper our aim is to draw attention to less visible aspects of antimicrobial stewardship (AMS) in residential aged care and their implications for nurse-led optimization of antibiotic use in these settings. By developing an account of the perceptions, experiences and practices of staff regarding the ‘on the ground’ work associated with implementing and upholding AMS objectives our study extends research on attempts to dismantle antibiotic infrastructures in Australian residential aged care facilities (RACF). Drawing on a review of relevant policies, empirical data is presented from fifty-six in-depth interviews conducted in 2021 with staff at 8 different RACFs. Interview participants included managers, nurses, and senior and junior personal care assistants. Our results suggest that registered nurses in residential aged care have been tasked with promoting antibiotic optimization and assigned with AMS responsibilities without sufficient authority and resourcing. A host of hidden care work associated with AMS strategies was evident, reinforcing some staff support for empirical antibiotic prescribing as a ‘safety net’ in uncertain clinical cases. We argue that this hidden work occurs where AMS strategies displace the infrastructural role previously performed by antibiotics, exposing structural gaps and pressures. The inability of organisational accounting systems and the broader AMS policy agenda to capture hidden AMS workflows in RACFs has consequences for future resourcing and organisational learning in ways that mean AMS gaps may remain unaddressed. These results support findings that AMS interventions might not be easily accepted by aged care staff in view of associated burdens which are under recognised and under supported in this domain.

An Isolated Genital Psoriasis: Uncommon Location That Make the Diagnostic Difficult

There are many dermatoses that can affect the genital area. These may be isolated genital conditions or signs of a general cutaneous disorder. Genital dermatoses can be classified as physiologic variants, and inflammatory, neoplastic, or infectious in etiology. Psoriasis is a chronic autoimmune disease, which has a prevalence of about 3% with genital involvement, reported in 29% a 40% of patients. Although the diagnosis of genital psoriasis can be aided by the presence of classic psoriatic plaques on the body, the penis may be the only affected region. In uncertain cases, a skin biopsy can lead to diagnosis. Initial management should consist of mild topical corticosteroids with increasing potency titrated to effect or topical calcineurin inhibitors. Systemic therapy with disease modifying small molecules or biologic drugs may be considered when disease is particularly bothersome, widespread, or unresponsive to conservative approache. This article provides a broad overview for dermatologists in addressing the Psoriasis wich is placed on. Recommendations for diagnostic evaluation, treatment, and appropriate follow-up are discussed.