The Impact of Opening a Medical Step-Down Unit on Medically Critically Ill Patient Outcomes and Throughput: A Difference-in-Differences Analysis.

OBJECTIVE: To understand the impact of adding a medical step-down unit (SDU) on patient outcomes and throughput in a medical intensive care unit (ICU). DESIGN: Retrospective cohort study. SETTING: Two academic tertiary care hospitals within the same health-care system. PATIENTS: Adults admitted to the medical ICU at either the control or intervention hospital from October 2013 to March 2014 (preintervention) and October 2014 to March 2015 (postintervention). INTERVENTIONS: Opening a 4-bed medical SDU at the intervention hospital on April 1, 2014. MEASUREMENTS AND MAIN RESULTS: Using standard summary statistics, we compared patients across hospitals. Using a difference-in-differences approach, we quantified the association of opening an SDU and outcomes (hospital mortality, hospital and ICU length of stay [LOS], and time to transfer to the ICU) after adjustment for secular trends in patient case-mix and patient-level covariates which might impact outcome. We analyzed 500 (245 pre- and 255 postintervention) patients in the intervention hospital and 678 (323 pre- and 355 postintervention) in the control hospital. Patients at the control hospital were younger (60.5-60.6 vs 64.0-65.4 years, P < .001) with a higher severity of acute illness at the time of evaluation for ICU admission (Sequential Organ Failure Assessment score: 4.9-4.0 vs 3.9-3.9, P < .001). Using the difference-in-differences methodology, we identified no association of hospital mortality (odds ratio [95% confidence interval]: 0.81 [0.42 to 1.55], P = .52) or hospital LOS (% change [95% confidence interval]: -8.7% [-28.6% to 11.2%], P = .39) with admission to the intervention hospital after SDU opening. The ICU LOS overall was not associated with admission to the intervention hospital in the postintervention period (-23.7% [-47.9% to 0.5%], P = .06); ICU LOS among survivors was significantly reduced (-27.5% [-50.5% to -4.6%], P = .019). Time to transfer to ICU was also significantly reduced (-26.7% [-44.7% to -8.8%], P = .004). CONCLUSIONS: Opening our medical SDU improved medical ICU throughput but did not affect more patient-centered outcomes of hospital mortality and LOS.

LCRNet: local cross-channel recalibration network for liver cancer classification based on CT images.

Liver cancer is the leading cause of mortality in the world. Over the years, researchers have spent much effort in developing computer-aided techniques to improve clinicians' diagnosis efficiency and precision, aiming at helping patients with liver cancer to take treatment early. Recently, attention mechanisms can enhance the representational power of convolutional neural networks (CNNs), which have been widely used in medical image analysis. In this paper, we propose a novel architectural unit, local cross-channel recalibration (LCR) module, dynamically adjusting the relative importance of intermediate feature maps by considering the roles of different global context features and building the local dependencies between channels. LCR first extracts different global context features and integrates them by global context integration operator, then estimates per channel attention weight with a local cross-channel interaction manner. We combine the LCR module with the residual block to form a Residual-LCR module and construct a deep neural network termed local cross-channel recalibration network (LCRNet) based on a stack of Residual-LCR modules to recognize live cancer atomically based on CT images. Furthermore, This paper collects a clinical CT image dataset of liver cancer, AMU-CT, to verify the effectiveness of LCRNet, which will be publicly available. The experiments on the AMU-CT dataset and public SD-OCT dataset demonstrate our LCRNet significantly outperforms state-of-the-art attention-based CNNs. Specifically, our LCRNet improves accuracy by over 11% than ECANet on the AMU-CT dataset. Supplementary Information: The online version contains supplementary material available at 10.1007/s13755-023-00263-6.

A deep learning approach to private data sharing of medical images using conditional generative adversarial networks (GANs).

Clinical data sharing can facilitate data-driven scientific research, allowing a broader range of questions to be addressed and thereby leading to greater understanding and innovation. However, sharing biomedical data can put sensitive personal information at risk. This is usually addressed by data anonymization, which is a slow and expensive process. An alternative to anonymization is construction of a synthetic dataset that behaves similar to the real clinical data but preserves patient privacy. As part of a collaboration between Novartis and the Oxford Big Data Institute, a synthetic dataset was generated based on images from COSENTYX® (secukinumab) ankylosing spondylitis (AS) clinical studies. An auxiliary classifier Generative Adversarial Network (ac-GAN) was trained to generate synthetic magnetic resonance images (MRIs) of vertebral units (VUs), conditioned on the VU location (cervical, thoracic and lumbar). Here, we present a method for generating a synthetic dataset and conduct an in-depth analysis on its properties along three key metrics: image fidelity, sample diversity and dataset privacy.

Mixed pyramid attention network for nuclear cataract classification based on anterior segment OCT images.

Nuclear cataract (NC) is a leading ocular disease globally for blindness and vision impairment. NC patients can improve their vision through cataract surgery or slow the opacity development with early intervention. Anterior segment optical coherence tomography (AS-OCT) image is an emerging ophthalmic image type, which can clearly observe the whole lens structure. Recently, clinicians have been increasingly studying the correlation between NC severity levels and clinical features from the nucleus region on AS-OCT images, and the results suggested the correlation is strong. However, automatic NC classification research based on AS-OCT images has rarely been studied. This paper presents a novel mixed pyramid attention network (MPANet) to classify NC severity levels on AS-OCT images automatically. In the MPANet, we design a novel mixed pyramid attention (MPA) block, which first applies the group convolution method to enhance the feature representation difference of feature maps and then construct a mixed pyramid pooling structure to extract local-global feature representations and different feature representation types simultaneously. We conduct extensive experiments on a clinical AS-OCT image dataset and a public OCT dataset to evaluate the effectiveness of our method. The results demonstrate that our method achieves competitive classification performance through comparisons to state-of-the-art methods and previous works. Moreover, this paper also uses the class activation mapping (CAM) technique to improve our method's interpretability of classification results.

Synthesis and Evaluation of LaBaCo2-xMoxO5+δ Cathode for Intermediate-Temperature Solid Oxide Fuel Cells.

LaBaCo2-xMoxO5+δ (LBCMx, x = 0-0.08) cathodes synthesized by a sol-gel method were evaluated for intermediate-temperature solid oxide fuel cells. The limit of the solid solubility of Mo in LBCMx was lower than 0.08. As the content of Mo increased gradually from 0 to 0.06, the thermal expansion coefficient decreased from 20.87 × 10-6 K-1 to 18.47 × 10-6 K-1. The introduction of Mo could increase the conductivity of LBCMx, which varied from 464 S cm-1 to 621 S cm-1 at 800 °C. The polarization resistance of the optimal cathode LBCM0.04 in air at 800 °C was 0.036 Ω cm2, reduced by a factor of 1.67 when compared with the undoped Mo cathode. The corresponding maximum power density of a single cell based on a YSZ electrolyte improved from 165 mW cm-2 to 248 mW cm-2 at 800 °C.

Supplemental Nutrition Assistance Program-Education Improves Food Security Independent of Food Assistance and Program Characteristics.

The purpose of this project was to determine whether consistent food assistance program participation or changes in participation over time mediated or moderated the effect of federal nutrition education through the Supplemental Nutrition Assistance Program-Education (SNAP-Ed) on food security and determine the associations of SNAP-Ed program delivery characteristics with change in food security. This secondary analysis used data from a randomized controlled trial from September 2013 through April 2015. SNAP-Ed-eligible participants (n = 328; ≥18 years) in households with children were recruited from 39 counties in Indiana, USA. The dependent variable was one year change in household food security score measured using the United States Household Food Security Survey Module. Assessment of mediation used Barron-Kenny analysis and moderation used interactions of food assistance program use and changes over time with treatment group in general linear regression modeling. Program delivery characteristics were investigated using mixed linear regression modeling. Results showed that neither consistent participation nor changes in food assistance program participation over time mediated nor moderated the effect of SNAP-Ed on food security and neither were SNAP-Ed program delivery characteristics associated with change in food security over the one year study period. SNAP-Ed directly improved food security among SNAP-Ed-eligible Indiana households with children regardless of food assistance program participation and changes over time or varying program delivery characteristics.