Search

Abstract Recently, the underlying mechanism for successful deep learning (DL) was presented based on a quantitative method that measures the quality of a single filter in each layer of a DL model, particularly VGG-16 trained on CIFAR-10. This method exemplifies that each filter identifies small clusters of possible output labels, with additional noise selected as labels outside the clusters. This feature is progressively sharpened with each layer, resulting in an enhanced signal-to-noise ratio (SNR), which leads to an increase in the accuracy of the DL network. In this study, this mechanism is verified for VGG-16 and EfficientNet-B0 trained on the CIFAR-100 and ImageNet datasets, and the main results are as follows. First, the accuracy and SNR progressively increase with the layers. Second, for a given deep architecture, the maximal error rate increases approximately linearly with the number of output labels. Third, similar trends were obtained for dataset labels in the range [3, 1000], thus supporting the universality of this mechanism. Understanding the performance of a single filter and its dominating features paves the way to highly dilute the deep architecture without affecting its overall accuracy, and this can be achieved by applying the filter’s cluster connections (AFCC).

Abstract Background Adults with sickle cell disease (SCD) suffer early mortality and high morbidity. Many are not affiliated with SCD centers, defined as no ambulatory visit with a SCD specialist in 2 years. Negative social determinants of health (SDOH) can impair access to care. Hypothesis: Negative SDOH are more likely to be experienced by unaffiliated adults than adults who regularly receive expert SCD care. Methods Cross-sectional analysis of the SCD Implementation Consortium (SCDIC) Registry, a convenience sample at 8 academic SCD centers in 2017–2019. A Distressed Communities Index (DCI) score was assigned to each registry member’s zip code. Insurance status and other barriers to care were self-reported. Most patients were enrolled in the clinic or hospital setting. Results The SCDIC Registry enrolled 288 Unaffiliated and 2110 Affiliated SCD patients, ages 15-45y. The highest DCI quintile accounted for 39% of both Unaffiliated and Affiliated patients. Lack of health insurance was reported by 19% of Unaffiliated versus 7% of Affiliated patients. The most frequently selected barriers to care for both groups were “previous bad experience with the healthcare system” (40%) and “Worry about Cost” (17%). SCD co-morbidities had no straightforward trend of association with Unaffiliated status. The 8 sites’ results varied. Conclusion The DCI economic measure of SDOH was not associated with Unaffiliated status of patients recruited in the health care delivery setting. SCDIC Registrants reside in more distressed communities than other Americans. Other SDOH themes of affordability and negative experiences might contribute to Unaffiliated status. Recruiting Unaffiliated SCD patients to care might benefit from systems adopting value-based patient-centered solutions.

Purpose: To identify clinical features which may predict the angle status of a large cohort of NVG eyes at the time of diagnosis. Observations: Chart review was performed for all NVG eyes from 2010 to 2022. Complete angle closure was defined as having >75 % PAS, partial angle closure as having 1–75 % PAS, and open angles as having 0 % PAS. Among 190 eyes (174 patients) with a diagnosis of NVG, 29 eyes (28 patients) had a prior NVG diagnosis and 32 eyes (31 patients) did not undergo gonioscopy; 129 eyes (115 patients, mean 65.5 years, 50 % women) had a gonioscopy documented at the time of diagnosis. There were 32 eyes (25 %) with open angles, 39 eyes (30 %) with partially closed angles, and 58 eyes (45 %) with completely closed angles. Mean BCVAs were 20/138 (logMar 0.84, CI = 0.78–0.90), 20/662 (logMar 1.52, CI = 1.41–1.62), and 20/4375 (logMar 2.34, CI = 2.17–2.51), respectively (p

lovo-cel (bb1111; LentiGlobin for sickle cell disease [SCD]) gene therapy (GT) comprises autologous transplantation of hematopoietic stem and progenitor cells transduced with the BB305 lentiviral vector encoding a modified β-globin gene (βA-T87Q ) to produce anti-sickling hemoglobin (HbAT87Q ). The efficacy and safety of lovo-cel for SCD are being evaluated in the ongoing phase 1/2 HGB-206 study (ClinicalTrials.gov: NCT02140554). The treatment process evolved over time, using learnings from outcomes in the initial patients to optimize lovo-cel's benefit-risk profile. Following modest expression of HbAT87Q in the initial patients (Group A, n = 7), alterations were made to the treatment process for patients subsequently enrolled in Group B (n = 2, patients B1 and B2), including improvements to cell collection and lovo-cel manufacturing. After 6 months, median Group A peripheral blood vector copy number (≥0.08 c/dg) and HbAT87Q levels (≥0.46 g/dL) were inadequate for substantial clinical effect but stable and sustained over 5.5 years; both markedly improved in Group B (patient B1: ≥0.53 c/dg and ≥2.69 g/dL; patient B2: ≥2.14 c/dg and ≥6.40 g/dL, respectively) and generated improved biologic and clinical efficacy in Group B, including higher total hemoglobin and decreased hemolysis. The safety of the lovo-cel for SCD treatment regimen largely reflected the known side effects of HSPC collection, busulfan conditioning regimen, and underlying SCD; acute myeloid leukemia was observed in two patients in Group A and deemed unlikely related to insertional oncogenesis. Changes made during development of the lovo-cel treatment process were associated with improved outcomes and provide lessons for future SCD GT studies.

Background Most adults with sickle cell disease will experience a silent cerebral infarction (SCI) or overt stroke. Identifying patient subgroups with increased stroke incidence is important for future clinical trials focused on stroke prevention. Our 3‐center prospective cohort study tested the primary hypothesis that adults with sickle cell disease and SCIs have a greater incidence of new stroke or SCI compared with those without SCI. A secondary aim focused on identifying additional risk factors for progressive infarcts, particularly traditional risk factors for stroke in adults. Methods and Results This observational study included adults with sickle cell disease and no history of stroke. Magnetic resonance imaging scans of the brain completed at baseline and >1 year later were reviewed by 3 radiologists for baseline SCIs and new or progressive infarcts on follow‐up magnetic resonance imaging. Stroke risk factors were abstracted from the medical chart. Time‐to‐event analysis was utilized for progressive infarcts. Median age was 24.1 years; 45.3% of 95 participants had SCIs on baseline magnetic resonance imaging. Progressive infarcts were present in 17 participants (17.9%), and the median follow‐up was 2.1 years. Incidence of new infarcts was 11.95 per 100 patient‐years (6.17–20.88) versus 3.74 per 100 patient‐years (1.21–8.73) in those with versus without prior SCI. Multivariable Cox regression showed that baseline SCI predicts progressive infarcts (hazard ratio, 3.46 [95% CI, 1.05–11.39]; P=0.041); baseline hypertension was also associated with progressive infarcts (hazard ratio, 3.23 [95% CI, 1.16–9.51]; P=0.025). Conclusions Selecting individuals with SCIs and hypertension for stroke prevention trials in sickle cell disease may enrich the study population with those at highest risk for infarct recurrence.

Agronomic crop production in California, including grains and animal feed, faces multiple challenges that could be addressed by extension support. This means that extension services need to understand priorities for using their limited resources. Until now, a statewide assessment of the most important topics for research and extension was lacking. A survey was conducted by the University of California Cooperative Extension (UCCE) to identify high-priority needs and inform extension programming based on grower, consultant, and allied industry input. Our goal was to compare the importance of different topics with the level of satisfaction regarding UCCE’s delivery of information on these topics. Survey respondents identified integrated pest management, nutrient and irrigation management, and variety testing as high-priority needs, with overall high satisfaction regarding UCCE’s program delivery on these topics. Topics needing more focus (high priority but below-average level of satisfaction) included testing new products, soil health management, and water conservation and storage. Areas of low priority and low satisfaction included niche marketing, emerging crops, organic production, harvest/post-harvest technology, salinity management, compost and manure management, and greenhouse gas emission reductions. To address stakeholder challenges, results from this study suggest that research and extension efforts should prioritize issues directly impacting on-farm crop production. At the same time, areas of low interest reflect a need for more support to engage farmers on these topics, particularly those concerning state environmental regulations and challenges to local and global food production and security.

Objective: We evaluated the impact of a telemedicine bridge clinic on treatment outcomes and cost for patients with opioid use disorder. Telemedicine bridge clinics deliver low-barrier rapid assessment of patients with opioid use disorder via audio-only and audiovisual telemedicine to facilitate induction on medication therapy and connection to ongoing care. Methods: A pre-post analysis of UPMC Health Plan member claims was performed to evaluate the impact of this intervention on the trajectory of care for patients with continuous coverage before and after bridge clinic visit(s). Results: Analysis included 150 UPMC Health Plan members evaluated at the bridge clinic between April 2020 and October 2021. At least one buprenorphine prescription was filled within 30 days by 91% of patients; median proportion of days covered by buprenorphine was 73.3%, 54.4%, and 50.6% at 30, 90, and 180 days after an initial visit compared to median of no buprenorphine claims 30 days prior among the same patients. Patients had an 18% decline in unplanned care utilization 30 days after initial Bridge Clinic visit, with a 62% reduction in unplanned care cost per member per month (PMPM), 38% reduction in medical cost PMPM, and 10% reduction in total PMPM (medical + pharmacy cost) at 180 days. Primary care, outpatient behavioral health, and laboratory costs increased while emergency department, urgent care, and inpatient costs declined. Conclusion: Utilization of a telemedicine bridge clinic was associated with buprenorphine initiation, linkage to ongoing care with retention including medication treatment, reduced unplanned care cost, and overall savings.

Purpose: Tube shunts can be inserted into the anterior chamber, ciliary sulcus, or pars plana. Sulcus tube placement can be challenging. This report demonstrates three techniques for guidewire-assisted sulcus tube insertion. Observations: The first technique uses a needle inserted through a paracentesis 180-degrees across from the tube entry site and creates an ab-interno sclerotomy through which the guidewire is inserted by docking it into the needle bevel. The second technique involves inserting the guidewire into the eye via a paracentesis and using microforceps to retrieve it through a sclerotomy. The third technique uses forceps to insert the guidewire into a paracentesis 180° across from the planned tube entry site and dock it into a needle bevel that has been inserted into the sulcus. Each of these techniques provides a reliable and reproducible way to insert a tube into the sulcus. Conclusions and importance: Guidewire-assisted tube entry offers a promising solution in cases of difficult sulcus tube placement without substantial additional cost.

Backround: The aim of this study is to evaluate how competent family medicine residents see themselves in the interventional competency goals targeted in emergency medicine rotation and the factors associated with their competency levels. Methods: The study is a methodological and cross-sectional study carried out between 01 May-31 October 2022. 138 family medicine residents who completed their emergency medicine rotation and agreed to participate in the study were included in the study. In the questionnaire used to collect data in the study, there were the Emergency Medicine Rotation Interventional Competence Scale (EMRICS) developed by the researchers based on the family medicine specialty training curriculum and sociodemographic characteristics. Results: The mean ages of the study group 29.71±4.27 (Mean±SD) years, and 60.10% (n:83) were women. The mean scores obtained from the scale in the study group were 59.86±11.72 (Mean±SD). While the interventional competence in which family physician residents had the highest proficiency was arterial blood gas with 97.80% (n:135), the lowest was tracheostomy with 15.20% (n:21). Evaluation of BLS (p:0.037), intubation (p:0.028), tracheostomy (p:0.034), trauma patient assessment (p:0.004) in the study group who applied five or more applications received higher EMRICS than those who did not apply at all. Conclusion: In order to achieve the desired competence in interventional competency goals, it is recommended to be repeated at least 5 times.

Objectives: To evaluate the diagnostic accuracy of carotid artery compression using a point-of-care ultrasound probe (POCUS-CAC) in reducing pulse check times and facilitating the detection of the return of spontaneous circulation (ROSC) during cardiopulmonary resuscitation (CPR) compared to manual palpation (MP). The secondary aim of the study is to assess the ability of POCUS-CAC to detect ROSC during ongoing chest compressions. Methods: This prospective study was conducted in a tertiary emergency department between January and June 2023. During CPR, POCUS-CAC was performed by placing a linear ultrasound probe transversely on the lateral neck to assess the compressibility of the carotid artery. Complete compression of the artery without any visible pulsation indicated no ROSC, while resistance to compression or partial compression suggested the presence of ROSC. Simultaneously, another clinician performed manual palpation of the femoral artery. The primary outcome assessed in this study was comparing ROSC detection between POCUS-CAC and traditional methods, and the secondary outcome was comparing the time taken to detect ROSC with each method, and the ability to detect ROSC during ongoing chest compressions. Results: The study included 41 cardiac arrest patients and analyzed 496 MP pulse and 1984 POCUS-CAC checks. The mean time to identify a pulse using POCUS-CAC was significantly shorter, at 2.3 (0.5–7.8, SD ± 1.2, 95% CI [2.25, 2.35]) s, compared to 4.7 (2.0–10.5, SD ± 1.8, 95% CI [4.54, 4.86]) s with MP (p = 0.004). Additionally, 52.9% of ROSC cases were detected earlier using POCUS-CAC, even during ongoing chest compressions. The sensitivity of POCUS-CAC was 100% (95% CI [80.5–100%]) and the specificity was 87.5% (95% CI [67.6–97.3%]). The POCUS-CAC method required less than 5 s in 99.996% of cases. Conclusions: POCUS-CAC significantly reduces pulse check times and enhances the early detection of ROSC during CPR, offering a reliable and rapid alternative to traditional manual palpation methods in emergency settings.

BackgroundGraduate medical education (GME) program leaders struggle to incorporate quality measures in the ambulatory care setting, leading to knowledge gaps on how to provide feedback to residents and programs. While nationally collected quality of care data are available, their reliability for individual resident learning and for GME program improvement is understudied.ObjectiveTo examine the reliability of the Healthcare Effectiveness Data and Information Set (HEDIS) clinical performance measures in family medicine and internal medicine GME programs and to determine whether HEDIS measures can inform residents and their programs with their quality of care.MethodsFrom 2014 to 2017, we collected HEDIS measures from 566 residents in 8 family medicine and internal medicine programs under one sponsoring institution. Intraclass correlation was performed to establish patient sample sizes required for 0.70 and 0.80 reliability levels at the resident and program levels. Differences between the patient sample sizes required for reliable measurement and the actual patients cared for by residents were calculated.ResultsThe highest reliability levels for residents (0.88) and programs (0.98) were found for the most frequently available HEDIS measure, colorectal cancer screening. At the GME program level, 87.5% of HEDIS measures had sufficient sample sizes for reliable measurement at alpha 0.7 and 75.0% at alpha 0.8. Most resident level measurements were found to be less reliable.ConclusionsGME programs may reliably evaluate HEDIS performance pooled at the program level, but less so at the resident level due to patient volume.

The current study examined associations of intersectional social identities on Black women's (N = 126) career self-efficacy and interests at a Historically Black College and University (HBCU). Structural models examined associations of different aspects of gender and racial identity on Social Cognitive Career Theory (SCCT) predictors (i.e., learning experiences and self-efficacy) for each RIASEC career interest. Social Cognitive Career Theory paths from learning experiences to career interests, via self-efficacy, were supported for all six career interests. For gender norms, domesticity directly predicted learning experiences and indirectly predicted interests for enterprising, investigative, social, and conventional themes; however, primacy of work conformity was not associated with learning experiences or indirect effects for any career interest. Racial centrality only predicted learning experiences and indirect effects on career interests for social careers. Aspects of racial and gender identity may set forth educational decisions that have implications for the eventual careers that many undergraduates pursue.