High-Throughput Deep Learning Detection of Mitral Regurgitation.

BACKGROUND: Diagnosis of mitral regurgitation (MR) requires careful evaluation by echocardiography with Doppler imaging. This study presents the development and validation of a fully automated deep learning pipeline for identifying apical 4-chamber view videos with color Doppler echocardiography and detecting clinically significant (moderate or severe) MR from transthoracic echocardiograms. METHODS: A total of 58 614 transthoracic echocardiograms (2 587 538 videos) from Cedars-Sinai Medical Center were used to develop and test an automated pipeline to identify apical 4-chamber view videos with color Doppler across the mitral valve and then assess MR severity. The model was tested internally on a test set of 1800 studies (80 833 videos) from Cedars-Sinai Medical Center and externally evaluated in a geographically distinct cohort of 915 studies (46 890 videos) from Stanford Healthcare. RESULTS: In the held-out Cedars-Sinai Medical Center test set, the view classifier demonstrated an area under the curve (AUC) of 0.998 (0.998-0.999) and correctly identified 3452 of 3539 echocardiography videos as having color Doppler information across the mitral valve (sensitivity of 0.975 [0.968-0.982] and specificity of 0.999 [0.999-0.999] compared with manually curated videos). In the external test cohort from Stanford Healthcare, the view classifier correctly identified 1051 of 1055 manually curated videos with color Doppler information across the mitral valve (sensitivity of 0.996 [0.990-1.000] and specificity of 0.999 [0.999-0.999]). In the Cedars-Sinai Medical Center test cohort, MR moderate or greater in severity was detected with an AUC of 0.916 (0.899-0.932) and severe MR was detected with an AUC of 0.934 (0.913-0.953). In the Stanford Healthcare test cohort, the model detected MR moderate or greater in severity with an AUC of 0.951 (0.924-0.973) and severe MR with an AUC of 0.969 (0.946-0.987). CONCLUSIONS: In this study, a novel automated pipeline for identifying clinically significant MR from full transthoracic echocardiography studies demonstrated excellent performance across large numbers of studies and across multiple institutions. Such an approach has the potential for automated screening and surveillance of MR.

Determinants of Mortality and Mid-Term Outcomes After Transcarotid Artery Revascularization and Transfemoral Carotid Artery Stenting.

OBJECTIVES: The potential benefit of transcarotid artery revascularization (TCAR) over transfemoral carotid artery stenting (tfCAS) has been studied in the perioperative period with lower rates of stroke and death; however, data on mid-term outcomes are limited. We aimed to evaluate 3-year outcomes after TCAR and tfCAS and determine the primary predictors of 30-day and 1-year mortality following TCAR. METHODS: Data from the Vascular Quality Initiative for patients undergoing TCAR or tfCAS from January 2016 to December 2022 were analyzed. 1:1 propensity score matching using the nearest-neighbor method was used to adjust baseline demographics and clinical characteristics. Kaplan-Meier survival analysis and Cox Proportional Hazard Regression were used to evaluate long-term outcomes. Iterative stepwise multiple logistic regression analysis and Cox Proportional Hazard Regression were used to identify predictors of 30-day and 1-year mortality, respectively, based upon preoperative, intraoperative, and postoperative factors. RESULTS: A total of 70 237 patients were included in analysis (TCAR=58.7%, tfCAS=41.3%). Transcarotid artery revascularization patients were older and had higher rates of comorbid conditions and high-risk medical and anatomic features than tfCAS patients. Propensity score matching yielded 22 322 pairs with no major differences between groups except that TCAR patients were older (71.6 years vs 70.8 years). At 3 years, TCAR was associated with a 24% reduction in hazard of death compared with tfCAS (hazard ratio [HR]=0.76, 95% confidence interval [CI]=0.71-0.82, p<0.001), for both symptomatic and asymptomatic patients. This survival advantage was established in the first 6 months (HR=0.59, 95% CI=0.53-0.62, p<0.001), with no difference in mortality risk from 6 months to 36 months (HR=0.95, 95% CI=0.86-1.05, p=0.31). Transcarotid artery revascularization was also associated with decreased hazard for 3-year stroke (HR=0.81, 95% CI=0.66-0.99, p=0.04) and stroke or death (HR=0.81, 95% CI=0.76-0.87, p<0.001) compared with tfCAS. The top predictors for 30-day and 1-year mortality were postoperative complications. The primary independent predictor was the occurrence of postoperative stroke. CONCLUSIONS: Transcarotid artery revascularization had a sustained mid-term survival advantage associated over tfCAS, with the benefit being established primarily within the first 6 months. Notably, our findings highlight the importance of postoperative stroke as the primary independent predictor for 30-day and 1-year mortal. CLINICAL IMPACT: The ongoing debate over the superiority of TCAR compared to tfCAS and CEA has been limited by a lack of comparative studies examining the impact of pre-operative symptoms on outcomes. Furthermore, data are scarce on mid-term outcomes for TCAR beyond the perioperative period. As a result, it remains uncertain whether the initial benefits of stroke and death reduction observed with TCAR over tfCAS persist beyond one year. Our study addresses these gaps in the literature, offering evidence to enable clinicians to assess the efficacy of TCAR for up to three years. Additionally, our study seeks to identify risk factors for postoperative mortality following TCAR, facilitating optimal patient stratification.

Molecular Interaction Between Butorphanol and κ-Opioid Receptor.

BACKGROUND: The misuse of opioids stems, in part, from inadequate knowledge of molecular interactions between opioids and opioid receptors. It is still unclear why some opioids are far more addictive than others. The κ-opioid receptor (KOR) plays a critical role in modulating pain, addiction, and many other physiological and pathological processes. Butorphanol, an opioid analgesic, is a less addictive opioid with unique pharmacological profiles. In this study, we investigated the interaction between butorphanol and KOR to obtain insights into the safe usage of this medication. METHODS: We determined the binding affinity of butorphanol to KOR with a naltrexone competition study. Recombinant KORs expressed in mammalian cell membranes (Chem-1) were used for G-protein activation studies, and a human embryonic kidney-293 (HEK-293) cell line stably transfected with the human KOR was used for ß-arrestin study as previously described in the literature. The effects of butorphanol on KOR internalization were investigated using mouse neuroblastoma Neuro2A cells stably transfected with mKOR-tdTomato fusion protein (N2A-mKOR-tdT) cells overexpressing KOR. The active-state KOR crystal structure was used for docking calculation of butorphanol to characterize the ligand binding site. Salvinorin A, a full KOR agonist, was used as a control for comparison. RESULTS: The affinity of KOR for butorphanol is characterized by Kd of 0.1 ± 0.02 nM, about 20-fold higher compared with that of the µ-opioid receptor (MOR; 2.4 ± 1.2 nM). Our data indicate that butorphanol is more potent on KOR than on MOR. In addition, butorphanol acts as a partial agonist of KOR in the G-protein activation pathway and is a full agonist on the ß-arrestin recruitment pathway, similar to that of salvinorin A. The activation of the ß-arrestin pathway is further confirmed by KOR internalization. The in silico docking model indicates that both salvinorin A and butorphanol share the same binding cavity with the KOR full agonist MP1104. This cavity plays an important role in determining either agonist or antagonist effects of the ligand. CONCLUSIONS: In conclusion, butorphanol is a partial KOR agonist in the G-protein activation pathway and a potent KOR full agonist in the ß-arrestin recruitment pathway. The structure analysis offers insights into the molecular mechanism of KOR interaction and activation by butorphanol.

pH Sensing Properties of Flexible, Bias-Free Graphene Microelectrodes in Complex Fluids: From Phosphate Buffer Solution to Human Serum.

Advances in techniques for monitoring pH in complex fluids can have a significant impact on analytical and biomedical applications. This study develops flexible graphene microelectrodes (GEs) for rapid (<5 s), very-low-power (femtowatt) detection of the pH of complex biofluids by measuring real-time Faradaic charge transfer between the GE and a solution at zero electrical bias. For an idealized sample of phosphate buffer solution (PBS), the Faradaic current is varied monotonically and systematically with the pH, with a resolution of ≈0.2 pH unit. The current-pH dependence is well described by a hybrid analytical-computational model, where the electric double layer derives from an intrinsic, pH-independent (positive) charge associated with the graphene-water interface and ionizable (negative) charged groups. For ferritin solution, the relative Faradaic current, defined as the difference between the measured current response and a baseline response due to PBS, shows a strong signal associated with ferritin disassembly and the release of ferric ions at pH ≈2.0. For samples of human serum, the Faradaic current shows a reproducible rapid (<20 s) response to pH. By combining the Faradaic current and real-time current variation, the methodology is potentially suitable for use to detect tumor-induced changes in extracellular pH.

Machine Learning and Bias in Medical Imaging: Opportunities and Challenges.

Bias in health care has been well documented and results in disparate and worsened outcomes for at-risk groups. Medical imaging plays a critical role in facilitating patient diagnoses but involves multiple sources of bias including factors related to access to imaging modalities, acquisition of images, and assessment (ie, interpretation) of imaging data. Machine learning (ML) applied to diagnostic imaging has demonstrated the potential to improve the quality of imaging-based diagnosis and the precision of measuring imaging-based traits. Algorithms can leverage subtle information not visible to the human eye to detect underdiagnosed conditions or derive new disease phenotypes by linking imaging features with clinical outcomes, all while mitigating cognitive bias in interpretation. Importantly, however, the application of ML to diagnostic imaging has the potential to either reduce or propagate bias. Understanding the potential gain as well as the potential risks requires an understanding of how and what ML models learn. Common risks of propagating bias can arise from unbalanced training, suboptimal architecture design or selection, and uneven application of models. Notwithstanding these risks, ML may yet be applied to improve gain from imaging across all 3A's (access, acquisition, and assessment) for all patients. In this review, we present a framework for understanding the balance of opportunities and challenges for minimizing bias in medical imaging, how ML may improve current approaches to imaging, and what specific design considerations should be made as part of efforts to maximize the quality of health care for all.

Impact of Case and Control Selection on Training Artificial Intelligence Screening of Cardiac Amyloidosis.

Background: Recent studies suggest that cardiac amyloidosis (CA) is significantly underdiagnosed. For rare diseases like CA, the optimal selection of cases and controls for artificial intelligence model training is unknown and can significantly impact model performance. Objectives: This study evaluates the performance of electrocardiogram (ECG) waveform-based artificial intelligence models for CA screening and assesses impact of different criteria for defining cases and controls. Methods: Using a primary cohort of ∼1.3 million ECGs from 341,989 patients, models were trained using different case and control definitions. Case definitions included ECGs from patients with an amyloidosis diagnosis by International Classification of Diseases-9/10 code, patients with CA, and patients seen in CA clinic. Models were then tested on test cohorts with identical selection criteria as well as a Cedars-Sinai general patient population cohort. Results: In matched held-out test data sets, different model AUCs ranged from 0.660 (95% CI: 0.642-0.736) to 0.898 (95% CI: 0.868-0.924). However, algorithms exhibited variable generalizability when tested on a Cedars-Sinai general patient population cohort, with AUCs dropping to 0.467 (95% CI: 0.443-0.491) to 0.898 (95% CI: 0.870-0.923). Models trained on more well-curated patient cases resulted in higher AUCs on similarly constructed test cohorts. However, all models performed similarly in the overall Cedars-Sinai general patient population cohort. A model trained with International Classification of Diseases 9/10 cases and population controls matched for age and sex resulted in the best screening performance. Conclusions: Models performed similarly in population screening, regardless of stringency of cases used during training, showing that institutions without dedicated amyloid clinics can train meaningful models on less curated CA cases. Additionally, AUC or other metrics alone are insufficient in evaluating deep learning algorithm performance. Instead, evaluation in the most clinically meaningful population is key.

Deep Learning Phenotyping of Tricuspid Regurgitation for Automated High Throughput Assessment of Transthoracic Echocardiography.

Background and Aims: Diagnosis of tricuspid regurgitation (TR) requires careful expert evaluation. This study developed an automated deep learning pipeline for assessing TR from transthoracic echocardiography. Methods: An automated deep learning workflow was developed using 47,312 studies (2,079,898 videos) from Cedars-Sinai Medical Center (CSMC) between 2011 and 2021. The pipeline was tested on a temporally distinct test set of 2,462 studies (108,138 videos) obtained in 2022 at CSMC and a geographically distinct cohort of 5,549 studies (278,377 videos) from Stanford Healthcare (SHC). Results: In the CSMC test dataset, the view classifier demonstrated an AUC of 1.000 (0.999 - 1.000) and identified at least one A4C video with colour Doppler across the tricuspid valve in 2,410 of 2,462 studies with a sensitivity of 0.975 (0.968-0.982) and a specificity of 1.000 (1.00-1.000). In the CSMC test cohort, moderate-or-severe TR was detected with an AUC of 0.928 (0.913 - 0.943) and severe TR was detected with an AUC of 0.956 (0.940 - 0.969). In the SHC cohort, the view classifier correctly identified at least one TR colour Doppler video in 5,268 of the 5,549 studies, resulting in an AUC of 0.999 (0.998 - 0.999), a sensitivity of 0.949 (0.944 - 0.955) and specificity of 0.999 (0.999 - 0.999). The AI model detected moderate-or-severe TR with an AUC of 0.951 (0.938 - 0.962) and severe TR with an AUC of 0.980 (0.966 - 0.988). Conclusions: We developed an automated pipeline to identify clinically significant TR with excellent performance. This approach carries potential for automated TR detection and stratification for surveillance and screening. Key Question: Can an automated deep learning model assess tricuspid regurgitation severity from echocardiography? Key Finding: We developed and validated an automated tricuspid regurgitation detection algorithm pipeline across two healthcare systems with high volume echocardiography labs. The algorithm correctly identifies apical-4-chamber view videos with colour Doppler across the tricuspid valve and grades clinically significant TR with strong agreement to expert clinical readers. Take Home message: A deep learning pipeline could automate TR screening, facilitating reproducible accurate assessment of TR severity, allowing rapid triage or re-review and expand access in low-resource or primary care settings.

Local aortic aneurysm wall expansion measured with automated image analysis.

BACKGROUND: Assessment of regional aortic wall deformation (RAWD) might better predict for abdominal aortic aneurysm (AAA) rupture than the maximal aortic diameter or growth rate. Using sequential computed tomography angiograms (CTAs), we developed a streamlined, semiautomated method of computing RAWD using deformable image registration (dirRAWD). METHODS: Paired sequential CTAs performed 1 to 2 years apart of 15 patients with AAAs of various shapes and sizes were selected. Using each patient's initial CTA, the luminal and aortic wall surfaces were segmented both manually and semiautomatically. Next, the same patient's follow-up CTA was aligned with the first using automated rigid image registration. Deformable image registration was then used to calculate the local aneurysm wall expansion between the sequential scans (dirRAWD). To measure technique accuracy, the deformable registration results were compared with the local displacement of anatomic landmarks (fiducial markers), such as the origin of the inferior mesenteric artery and/or aortic wall calcifications. Additionally, for each patient, the maximal RAWD was manually measured for each aneurysm and was compared with the dirRAWD at the same location. RESULTS: The technique was successful in all patients. The mean landmark displacement error was 0.59 ± 0.93 mm with no difference between true landmark displacement and deformable registration landmark displacement by Wilcoxon rank sum test (P = .39). The absolute difference between the manually measured maximal RAWD and dirRAWD was 0.27 ± 0.23 mm, with a relative difference of 7.9% and no difference using the Wilcoxon rank sum test (P = .69). No differences were found in the maximal dirRAWD when derived using a purely manual AAA segmentation compared with using semiautomated AAA segmentation (P = .55). CONCLUSIONS: We found accurate and automated RAWD measurements were feasible with clinically insignificant errors. Using semiautomated AAA segmentations for deformable image registration methods did not alter maximal dirRAWD accuracy compared with using manual AAA segmentations. Future work will compare dirRAWD with finite element analysis-derived regional wall stress and determine whether dirRAWD might serve as an independent predictor of rupture risk.

Middle School Students Effectively Improve Stroke Knowledge and Pass Them to Family Members in China Using Stroke 1-2-0.

Background and Purpose: This study tests the hypothesis that middle school and high school students can improve their stroke knowledge using Stroke 1-2-0, a stroke educational tool, and pass this knowledge on to their family members. Methods: A total of 625 students and 198 parents/grandparents participated in learning about stroke using Stroke 1-2-0. After a group training session for the students by a neurologist at school, the students took educational material to home and educated their parents/grandparents. A questionnaire was given to students, parents/grandparents before, immediately after, and 1 year after the educational event. Results: All participants agreed that Stroke 1-2-0 was a much easier tool to remember than FAST. Almost all the students (96.4%) remembered the meaning of Stroke 1-2-0 as compared to 7.3% from the base line (p < 0.001). The rate of complete Stroke 1-2-0 mastery from 96.3% fell to 84.4% at 3 months and 63.8% at 1 year after training (p < 0.001). Following education from children, the proportion of parents/grandparents who mastered Stroke 1-2-0 was significantly higher than baseline (79.9 vs. 24.8%). Conclusion: Middle school and high school students can effectively use Stroke 1-2-0 to improve their stroke knowledge and pass this knowledge to their family members. Sustained educational efforts and repeated educational events are needed though.

Too Young to Have a Stroke?-a Global Health Crisis.

This editorial discusses the importance of improving awareness of stroke in young individuals. Stoke can occur in any age group and is not restricted to elderly populations. Today, the average age of the first-time stroke patient continues to decrease. However, the incidence of stroke in seemingly healthy, young adults remains neglected, and stroke awareness among young patients remains poor, even in well-developed countries. Education targeting two common barriers to stroke care, identification and rescue, should be implemented for both medical professionals and the public domain. Only through education can we reduce preventable stroke-related death and damage in young patients moving forward.

Scalable Production of High-Sensitivity, Label-Free DNA Biosensors Based on Back-Gated Graphene Field Effect Transistors.

Scalable production of all-electronic DNA biosensors with high sensitivity and selectivity is a critical enabling step for research and applications associated with detection of DNA hybridization. We have developed a scalable and very reproducible (>90% yield) fabrication process for label-free DNA biosensors based upon graphene field effect transistors (GFETs) functionalized with single-stranded probe DNA. The shift of the GFET sensor Dirac point voltage varied systematically with the concentration of target DNA. The biosensors demonstrated a broad analytical range and limit of detection of 1 fM for 60-mer DNA oligonucleotide. In control experiments with mismatched DNA oligomers, the impact of the mismatch position on the DNA hybridization strength was confirmed. This class of highly sensitive DNA biosensors offers the prospect of detection of DNA hybridization and sequencing in a rapid, inexpensive, and accurate way.