National cross-sectional study on cost consciousness, cost accuracy, and national medical waste reduction initiative knowledge among pediatric hospitalists in the United States.

BACKGROUND/OBJECTIVE: Despite initiatives to reduce waste and spending, there is a gap in physician knowledge regarding the cost of commonly ordered items. We examined the relationship between pediatric hospitalists' knowledge of national medical waste reduction initiatives, self-reported level of cost-consciousness (the degree in which cost affects practice), and cost accuracy (how close an estimate is to its hospital cost) at a national level. METHODS: This cross-sectional study used a national, online survey sent to hospitalists at 49 children's hospitals to assess their knowledge of national medical waste reduction initiatives, self-reported cost consciousness, and cost estimates for commonly ordered laboratory studies, medications, and imaging studies. Actual unit costs for each hospital were obtained from the Pediatric Health Information System (PHIS). Cost accuracy was calculated as the percent difference between each respondent's estimate and unit costs, using cost-charge ratios (CCR). RESULTS: The hospitalist response rate was 17.7% (327/1850), representing 40 hospitals. Overall, 33.1% of respondents had no knowledge of national medical waste reduction initiatives and 24.3% had no knowledge of local hospital costs. There was no significant relationship between cost accuracy and knowledge of national medical waste reduction initiatives or high self-reported cost consciousness. Hospitalists with the highest self-reported cost consciousness were the least accurate in estimating costs for commonly ordered laboratory studies, medications, or imaging studies. Respondents overestimated the cost of all items with the largest percent difference with medications. Hospitalists practicing over 15 years had the highest cost accuracy. CONCLUSIONS: A large proportion of pediatric hospitalists lack knowledge on national waste reduction initiatives. Improving the cost-accuracy of pediatric hospitalists may not reduce health care costs as they overestimated many hospital costs. Median unit cost lists could be a resource for educating medical students and residents about health care costs.

Mortality impact of low CAC density predominantly occurs in early atherosclerosis: explainable ML in the CAC consortium.

BACKGROUND: Machine learning (ML) models of risk prediction with coronary artery calcium (CAC) and CAC characteristics exhibit high performance, but are not inherently interpretable. OBJECTIVES: To determine the direction and magnitude of impact of CAC characteristics on 10-year all-cause mortality (ACM) with explainable ML. METHODS: We analyzed asymptomatic subjects in the CAC consortium. We trained ML models on 80% and tested on 20% of the data with XGBoost, using clinical characteristics â€‹+ â€‹CAC (ML 1) and additional CAC characteristics of CAC density and number of calcified vessels (ML 2). We applied SHAP, an explainable ML tool, to explore the relationship of CAC and CAC characteristics with 10-year all-cause and CV mortality. RESULTS: 2376 deaths occurred among 63,215 patients [68% male, median age 54 (IQR 47-61), CAC 3 (IQR 0-94.3)]. ML2 was similar to ML1 to predict all-cause mortality (Area Under the Curve (AUC) 0.819 vs 0.821, p â€‹= â€‹0.23), but superior for CV mortality (0.847 vs 0.845, p â€‹= â€‹0.03). Low CAC density increased mortality impact, particularly ≤0.75. Very low CAC density ≤0.75 was present in only 4.3% of the patients with measurable density, and 75% occurred in CAC1-100. The number of diseased vessels did not increase mortality overall when simultaneously accounting for CAC and CAC density. CONCLUSION: CAC density contributes to mortality risk primarily when it is very low ≤0.75, which is primarily observed in CAC 1-100. CAC and CAC density are more important for mortality prediction than the number of diseased vessels, and improve prediction of CV but not all-cause mortality. Explainable ML techniques are useful to describe granular relationships in otherwise opaque prediction models.

Trends in Pediatric Rhabdomyolysis and Associated Renal Failure: A 10-Year Population-Based Study.

OBJECTIVES: Rhabdomyolysis in children is a highly variable condition with presentations ranging from myalgias to more severe complications like acute renal failure. We sought to explore demographics and incidence of pediatric rhabdomyolysis hospitalizations and rates of associated renal failure, as our current understanding is limited. METHODS: This was a retrospective analysis using the Healthcare Cost and Utilization Project Kids' Inpatient Database to identify children hospitalized with a primary diagnosis of rhabdomyolysis. Data were analyzed for demographic characteristics, as well as geographic and temporal trends. Multivariable logistic regression was used to identify characteristics associated with rhabdomyolysis-associated acute renal failure. RESULTS: From 2006 to 2016, there were 8599 hospitalized children with a primary diagnosis of rhabdomyolysis. Overall, hospitalizations for pediatric rhabdomyolysis are increasing over time, with geographic peaks in the South and Northeast regions, and seasonal peaks in March and August. Though renal morbidity was diagnosed in 8.5% of children requiring hospitalization for rhabdomyolysis, very few of these patients required renal replacement therapy (0.41%), and death was rare (0.03%). Characteristics associated with renal failure included male sex, age greater than 15 years, and non-Hispanic Black race. CONCLUSIONS: Though renal failure occurs at a significant rate in children hospitalized with rhabdomyolysis, severe complications, including death, are rare. The number of children hospitalized with rhabdomyolysis varies by geographic region and month of the year. Future studies are needed to explore etiologies of rhabdomyolysis and laboratory values that predict higher risk of morbidity and mortality in children with rhabdomyolysis.

Hospital-Level Neighborhood Opportunity and Rehospitalization for Common Diagnoses at US Children's Hospitals.

OBJECTIVE: Neighborhood conditions influence child health outcomes, but data examining association between local factors and hospital utilization are lacking. We determined if hospitals' mix of patients by neighborhood opportunity correlates with rehospitalization for common diagnoses at US children's hospitals. METHODS: We analyzed all discharges in 2018 for children ≤18 years at 47 children's hospitals for 14 common diagnoses. The exposure was hospital-level mean neighborhood opportunity - measured by Child Opportunity Index (COI) - for each diagnosis. The outcome was same-cause rehospitalization within 365 days. We measured association via Pearson correlation coefficient. For diagnoses with significant associations, we also examined shorter rehospitalization time windows and compared unadjusted and COI-adjusted rehospitalization rates. RESULTS: There were 256,871 discharges included. Hospital-level COI ranged from 17th to 70th percentile nationally. Hospitals serving lower COI neighborhoods had more frequent rehospitalization for asthma (ρ -0.34 [95% confidence interval -0.57, -0.06]) and diabetes (ρ -0.33 [-0.56, -0.04]), but fewer primary mental health rehospitalizations (ρ 0.47 [0.21, 0.67]). There was no association for 11 other diagnoses. Secondary timepoint analysis revealed increasing correlation over time, with differences by diagnosis. Adjustment for hospital-level COI resulted in 26%, 32%, and 45% of hospitals changing >1 decile in rehospitalization rank order for diabetes, asthma, and mental health diagnoses, respectively. CONCLUSIONS: Children's hospitals vary widely in their mix of neighborhoods served. Asthma, diabetes, and mental health rehospitalization rates correlate with COI, suggesting that neighborhood factors may influence outcome disparities for these conditions. Hospital outcomes may be affected by neighborhood opportunity, which has implications for benchmarking.

Current and Future Applications of Artificial Intelligence in Coronary Artery Disease.

Cardiovascular diseases (CVDs) carry significant morbidity and mortality and are associated with substantial economic burden on healthcare systems around the world. Coronary artery disease, as one disease entity under the CVDs umbrella, had a prevalence of 7.2% among adults in the United States and incurred a financial burden of 360 billion US dollars in the years 2016-2017. The introduction of artificial intelligence (AI) and machine learning over the last two decades has unlocked new dimensions in the field of cardiovascular medicine. From automatic interpretations of heart rhythm disorders via smartwatches, to assisting in complex decision-making, AI has quickly expanded its realms in medicine and has demonstrated itself as a promising tool in helping clinicians guide treatment decisions. Understanding complex genetic interactions and developing clinical risk prediction models, advanced cardiac imaging, and improving mortality outcomes are just a few areas where AI has been applied in the domain of coronary artery disease. Through this review, we sought to summarize the advances in AI relating to coronary artery disease, current limitations, and future perspectives.

Plaque Character and Progression According to the Location of Coronary Atherosclerotic Plaque.

Although acute coronary syndrome culprit lesions occur more frequently in the proximal coronary artery, whether the proximal clustering of high-risk plaque is reflected in earlier-stage atherosclerosis remains unclarified. We evaluated the longitudinal distribution of stable atherosclerotic lesions on coronary computed tomography angiography (CCTA) in 1,478 patients (mean age, 61 years; men, 58%) enrolled from a prospective multinational registry of consecutive patients undergoing serial CCTA. Of 3,202 coronary artery lesions identified, 2,140 left lesions were classified (based on the minimal lumen diameter location) into left main (LM, n = 128), proximal (n = 739), and other (n = 1,273), and 1,062 right lesions were classified into proximal (n = 355) and other (n = 707). Plaque volume (PV) was the highest in proximal lesions (median, 26.1 mm3), followed by LM (20.6 mm3) and other lesions (15.0 mm3, p <0.001), for left lesions, and was lager in proximal (25.8 mm3) than in other lesions (15.2 mm3, p <0.001) for right lesions. On both sides, proximally located lesions tended to have greater necrotic core and fibrofatty components than other lesions (left: LM, 10.6%; proximal, 5.8%; other, 3.4% of the total PV, p <0.001; right: proximal, 8.4%; other 3.1%, p <0.001), with less calcified plaque component (left: LM, 18.3%; proximal, 30.3%; other, 37.7%, p <0.001; right: proximal, 23.3%, other, 36.6%, p <0.001), and tended to progress rapidly (adjusted odds ratios: left: LM, reference; proximal, 0.95, p = 0.803; other, 0.64, p = 0.017; right: proximal, reference; other, 0.52, p <0.001). Proximally located plaques were larger, with more risky composition, and progressed more rapidly.

Differential progression of coronary atherosclerosis according to plaque composition: a cluster analysis of PARADIGM registry data.

Patient-specific phenotyping of coronary atherosclerosis would facilitate personalized risk assessment and preventive treatment. We explored whether unsupervised cluster analysis can categorize patients with coronary atherosclerosis according to their plaque composition, and determined how these differing plaque composition profiles impact plaque progression. Patients with coronary atherosclerotic plaque (n = 947; median age, 62 years; 59% male) were enrolled from a prospective multi-national registry of consecutive patients who underwent serial coronary computed tomography angiography (median inter-scan duration, 3.3 years). K-means clustering applied to the percent volume of each plaque component and identified 4 clusters of patients with distinct plaque composition. Cluster 1 (n = 52), which comprised mainly fibro-fatty plaque with a significant necrotic core (median, 55.7% and 16.0% of the total plaque volume, respectively), showed the least total plaque volume (PV) progression (+ 23.3 mm3), with necrotic core and fibro-fatty PV regression (- 5.7 mm3 and - 5.6 mm3, respectively). Cluster 2 (n = 219), which contained largely fibro-fatty (39.2%) and fibrous plaque (46.8%), showed fibro-fatty PV regression (- 2.4 mm3). Cluster 3 (n = 376), which comprised mostly fibrous (62.7%) and calcified plaque (23.6%), showed increasingly prominent calcified PV progression (+ 21.4 mm3). Cluster 4 (n = 300), which comprised mostly calcified plaque (58.7%), demonstrated the greatest total PV increase (+ 50.7mm3), predominantly increasing in calcified PV (+ 35.9 mm3). Multivariable analysis showed higher risk for plaque progression in Clusters 3 and 4, and higher risk for adverse cardiac events in Clusters 2, 3, and 4 compared to that in Cluster 1. Unsupervised clustering algorithms may uniquely characterize patient phenotypes with varied atherosclerotic plaque profiles, yielding distinct patterns of progressive disease and outcome.

Prevalence and Epidemiology of Mycoplasma genitalium in a Pacific-Region Military Population.

BACKGROUND: Mycoplasma genitalium is an important emerging sexually transmitted pathogen commonly causing urethritis in men, cervicitis, and pelvic inflammatory disease in women with potential of infertility. Accumulating evidence identifies the prevalence of M. genitalium similar to long recognized pathogens, Chlamydia trachomatis and Neisseria gonorrhoeae. The purpose of this study was to establish the prevalence and epidemiology of M. genitalium in a mid-Pacific military population. METHODS: A prospective analysis was conducted from routine specimens collected as standard of care for sexually transmitted infection (STI) testing at Tripler Army Medical Center on Oahu, HI. The prevalence of M. genitalium was determined using the Aptima M. genitalium assay, a transcription-mediated amplification test. A multivariate analysis was performed to assess the associations for this infection with other STIs and demographic factors. RESULTS: A total of 1876 specimens were tested in a 6-month period including 6 sample types from 1158 females and 718 males. Subject ages ranged from 18 to 76 years, with a median of 24 years (interquartile range, 21-29 years). The prevalence of M. genitalium was 8.8% overall (n = 165), 7.1% in females and 11.6% in males. Coinfection with M. genitalium occurred with another sexually-transmitted pathogen in 43 patients (18.3%), with C. trachomatis as the most common organism (n = 38). CONCLUSIONS: These data contribute to the evidence base for M. genitalium and STI screening in an active-duty military.

End-to-End, Pixel-Wise Vessel-Specific Coronary and Aortic Calcium Detection and Scoring Using Deep Learning.

Conventional scoring and identification methods for coronary artery calcium (CAC) and aortic calcium (AC) result in information loss from the original image and can be time-consuming. In this study, we sought to demonstrate an end-to-end deep learning model as an alternative to the conventional methods. Scans of 377 patients with no history of coronary artery disease (CAD) were obtained and annotated. A deep learning model was trained, tested and validated in a 60:20:20 split. Within the cohort, mean age was 64.2 ± 9.8 years, and 33% were female. Left anterior descending, right coronary artery, left circumflex, triple vessel, and aortic calcifications were present in 74.87%, 55.82%, 57.41%, 46.03%, and 85.41% of patients respectively. An overall Dice score of 0.952 (interquartile range 0.921, 0.981) was achieved. Stratified by subgroups, there was no difference between male (0.948, interquartile range 0.920, 0.981) and female (0.965, interquartile range 0.933, 0.980) patients (p = 0.350), or, between age <65 (0.950, interquartile range 0.913, 0.981) and age ≥65 (0.957, interquartile range 0.930, 0.9778) (p = 0.742). There was good correlation and agreement for CAC prediction (rho = 0.876, p < 0.001), with a mean difference of 11.2% (p = 0.100). AC correlated well (rho = 0.947, p < 0.001), with a mean difference of 9% (p = 0.070). Automated segmentation took approximately 4 s per patient. Taken together, the deep-end learning model was able to robustly identify vessel-specific CAC and AC with high accuracy, and predict Agatston scores that correlated well with manual annotation, facilitating application into areas of research and clinical importance.

Comparing a novel machine learning method to the Friedewald formula and Martin-Hopkins equation for low-density lipoprotein estimation.

BACKGROUND: Low-density lipoprotein cholesterol (LDL-C) is a target for cardiovascular prevention. Contemporary equations for LDL-C estimation have limited accuracy in certain scenarios (high triglycerides [TG], very low LDL-C). OBJECTIVES: We derived a novel method for LDL-C estimation from the standard lipid profile using a machine learning (ML) approach utilizing random forests (the Weill Cornell model). We compared its correlation to direct LDL-C with the Friedewald and Martin-Hopkins equations for LDL-C estimation. METHODS: The study cohort comprised a convenience sample of standard lipid profile measurements (with the directly measured components of total cholesterol [TC], high-density lipoprotein cholesterol [HDL-C], and TG) as well as chemical-based direct LDL-C performed on the same day at the New York-Presbyterian Hospital/Weill Cornell Medicine (NYP-WCM). Subsequently, an ML algorithm was used to construct a model for LDL-C estimation. Results are reported on the held-out test set, with correlation coefficients and absolute residuals used to assess model performance. RESULTS: Between 2005 and 2019, there were 17,500 lipid profiles performed on 10,936 unique individuals (4,456 females; 40.8%) aged 1 to 103. Correlation coefficients between estimated and measured LDL-C values were 0.982 for the Weill Cornell model, compared to 0.950 for Friedewald and 0.962 for the Martin-Hopkins method. The Weill Cornell model was consistently better across subgroups stratified by LDL-C and TG values, including TG >500 and LDL-C <70. CONCLUSIONS: An ML model was found to have a better correlation with direct LDL-C than either the Friedewald formula or Martin-Hopkins equation, including in the setting of elevated TG and very low LDL-C.

A Boosted Ensemble Algorithm for Determination of Plaque Stability in High-Risk Patients on Coronary CTA.

OBJECTIVES: This study sought to identify culprit lesion (CL) precursors among acute coronary syndrome (ACS) patients based on qualitative and quantitative computed tomography-based plaque characteristics. BACKGROUND: Coronary computed tomography angiography (CTA) has been validated for patient-level prediction of ACS. However, the applicability of coronary CTA to CL assessment is not known. METHODS: Utilizing the ICONIC (Incident COroNary Syndromes Identified by Computed Tomography) study, a nested case-control study of 468 patients with baseline coronary CTA, the study included ACS patients with invasive coronary angiography-adjudicated CLs that could be aligned to CL precursors on baseline coronary CTA. Separate blinded core laboratories adjudicated CLs and performed atherosclerotic plaque evaluation. Thereafter, the study used a boosted ensemble algorithm (XGBoost) to develop a predictive model of CLs. Data were randomly split into a training set (80%) and a test set (20%). The area under the receiver-operating characteristic curve of this model was compared with that of diameter stenosis (model 1), high-risk plaque features (model 2), and lesion-level features of CL precursors from the ICONIC study (model 3). Thereafter, the machine learning (ML) model was applied to 234 non-ACS patients with 864 lesions to determine model performance for CL exclusion. RESULTS: CL precursors were identified by both coronary angiography and baseline coronary CTA in 124 of 234 (53.0%) patients, with a total of 582 lesions (containing 124 CLs) included in the analysis. The ML model demonstrated significantly higher area under the receiver-operating characteristic curve for discriminating CL precursors (0.774; 95% confidence interval [CI]: 0.758 to 0.790) compared with model 1 (0.599; 95% CI: 0.599 to 0.599; p < 0.01), model 2 (0.532; 95% CI: 0.501 to 0.563; p < 0.01), and model 3 (0.672; 95% CI: 0.662 to 0.682; p < 0.01). When applied to the non-ACS cohort, the ML model had a specificity of 89.3% for excluding CLs. CONCLUSIONS: In a high-risk cohort, a boosted ensemble algorithm can be used to predict CL from non-CL precursors on coronary CTA.

Extraction of radiographic findings from unstructured thoracoabdominal computed tomography reports using convolutional neural network based natural language processing.

BACKGROUND: Heart failure (HF) is a major cause of morbidity and mortality. However, much of the clinical data is unstructured in the form of radiology reports, while the process of data collection and curation is arduous and time-consuming. PURPOSE: We utilized a machine learning (ML)-based natural language processing (NLP) approach to extract clinical terms from unstructured radiology reports. Additionally, we investigate the prognostic value of the extracted data in predicting all-cause mortality (ACM) in HF patients. MATERIALS AND METHODS: This observational cohort study utilized 122,025 thoracoabdominal computed tomography (CT) reports from 11,808 HF patients obtained between 2008 and 2018. 1,560 CT reports were manually annotated for the presence or absence of 14 radiographic findings, in addition to age and gender. Thereafter, a Convolutional Neural Network (CNN) was trained, validated and tested to determine the presence or absence of these features. Further, the ability of CNN to predict ACM was evaluated using Cox regression analysis on the extracted features. RESULTS: 11,808 CT reports were analyzed from 11,808 patients (mean age 72.8 ± 14.8 years; 52.7% (6,217/11,808) male) from whom 3,107 died during the 10.6-year follow-up. The CNN demonstrated excellent accuracy for retrieval of the 14 radiographic findings with area-under-the-curve (AUC) ranging between 0.83-1.00 (F1 score 0.84-0.97). Cox model showed the time-dependent AUC for predicting ACM was 0.747 (95% confidence interval [CI] of 0.704-0.790) at 30 days. CONCLUSION: An ML-based NLP approach to unstructured CT reports demonstrates excellent accuracy for the extraction of predetermined radiographic findings, and provides prognostic value in HF patients.

Machine learning insight into the role of imaging and clinical variables for the prediction of obstructive coronary artery disease and revascularization: An exploratory analysis of the CONSERVE study.

BACKGROUND: Machine learning (ML) is able to extract patterns and develop algorithms to construct data-driven models. We use ML models to gain insight into the relative importance of variables to predict obstructive coronary artery disease (CAD) using the Coronary Computed Tomographic Angiography for Selective Cardiac Catheterization (CONSERVE) study, as well as to compare prediction of obstructive CAD to the CAD consortium clinical score (CAD2). We further perform ML analysis to gain insight into the role of imaging and clinical variables for revascularization. METHODS: For prediction of obstructive CAD, the entire ICA arm of the study, comprising 719 patients was used. For revascularization, 1,028 patients were randomized to invasive coronary angiography (ICA) or coronary computed tomographic angiography (CCTA). Data was randomly split into 80% training 20% test sets for building and validation. Models used extreme gradient boosting (XGBoost). RESULTS: Mean age was 60.6 ± 11.5 years and 64.3% were female. For the prediction of obstructive CAD, the AUC was significantly higher for ML at 0.779 (95% CI: 0.672-0.886) than for CAD2 (0.696 [95% CI: 0.594-0.798]) (P = 0.01). BMI, age, and angina severity were the most important variables. For revascularization, the model obtained an overall area under the receiver-operation curve (AUC) of 0.958 (95% CI = 0.933-0.983). Performance did not differ whether the imaging parameters used were from ICA (AUC 0.947, 95% CI = 0.903-0.990) or CCTA (AUC 0.941, 95% CI = 0.895-0.988) (P = 0.90). The ML model obtained sensitivity and specificity of 89.2% and 92.9%, respectively. Number of vessels with ≥70% stenosis, maximum segment stenosis severity (SSS) and body mass index (BMI) were the most important variables. Exclusion of imaging variables resulted in performance deterioration, with an AUC of 0.705 (95% CI 0.614-0.795) (P <0.0001). CONCLUSIONS: For obstructive CAD, the ML model outperformed CAD2. BMI is an important variable, although currently not included in most scores. In this ML model, imaging variables were most associated with revascularization. Imaging modality did not influence model performance. Removal of imaging variables reduced model performance.

Automatic segmentation of multiple cardiovascular structures from cardiac computed tomography angiography images using deep learning.

OBJECTIVES: To develop, demonstrate and evaluate an automated deep learning method for multiple cardiovascular structure segmentation. BACKGROUND: Segmentation of cardiovascular images is resource-intensive. We design an automated deep learning method for the segmentation of multiple structures from Coronary Computed Tomography Angiography (CCTA) images. METHODS: Images from a multicenter registry of patients that underwent clinically-indicated CCTA were used. The proximal ascending and descending aorta (PAA, DA), superior and inferior vena cavae (SVC, IVC), pulmonary artery (PA), coronary sinus (CS), right ventricular wall (RVW) and left atrial wall (LAW) were annotated as ground truth. The U-net-derived deep learning model was trained, validated and tested in a 70:20:10 split. RESULTS: The dataset comprised 206 patients, with 5.130 billion pixels. Mean age was 59.9 ± 9.4 yrs., and was 42.7% female. An overall median Dice score of 0.820 (0.782, 0.843) was achieved. Median Dice scores for PAA, DA, SVC, IVC, PA, CS, RVW and LAW were 0.969 (0.979, 0.988), 0.953 (0.955, 0.983), 0.937 (0.934, 0.965), 0.903 (0.897, 0.948), 0.775 (0.724, 0.925), 0.720 (0.642, 0.809), 0.685 (0.631, 0.761) and 0.625 (0.596, 0.749) respectively. Apart from the CS, there were no significant differences in performance between sexes or age groups. CONCLUSIONS: An automated deep learning model demonstrated segmentation of multiple cardiovascular structures from CCTA images with reasonable overall accuracy when evaluated on a pixel level.

Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure.

Background: Existing risk assessment tools for heart failure (HF) outcomes use structured databases with static, single-timepoint clinical data and have limited accuracy. Objective: The purpose of this study was to develop a comprehensive approach for accurate prediction of 30-day unplanned readmission and all-cause mortality (ACM) that integrates clinical and physiological data available in the electronic health record system. Methods: Three predictive models for 30-day unplanned readmissions or ACM were created using an extreme gradient boosting approach: (1) index admission model; (2) index discharge model; and (3) feature-aggregated model. Performance was assessed by the area under the curve (AUC) metric and compared with that of the HOSPITAL score, a widely used predictive model for hospital readmission. Results: A total of 3774 patients with a primary billing diagnosis of HF were included (614 experienced the primary outcome), with 796 variables used in the admission and discharge models, and 2032 in the feature-aggregated model. The index admission model had AUC = 0.723, the index discharge model had AUC = 0.754, and the feature-aggregated model had AUC = 0.756 for prediction of 30-day unplanned readmission or ACM. For comparison, the HOSPITAL score had AUC = 0.666 (admission model: P = .093; discharge model: P = .022; feature aggregated: P = .012). Conclusion: These models predict risk of HF hospitalizations and ACM in patients admitted with HF and emphasize the importance of incorporating large numbers of variables in machine learning models to identify predictors for future investigation.

Automated dynamic motion correction using normalized gradient fields for 82rubidium PET myocardial blood flow quantification.

BACKGROUND: Patient motion can lead to misalignment of left ventricular (LV) volumes-of-interest (VOIs) and subsequently inaccurate quantification of myocardial blood flow (MBF) and flow reserve (MFR) from dynamic PET myocardial perfusion images. We aimed to develop an image-based 3D-automated motion-correction algorithm that corrects the full dynamic sequence for translational motion, especially in the early blood phase frames (~ first minute) where the injected tracer activity is transitioning from the blood pool to the myocardium and where conventional image registration algorithms have had limited success. METHODS: We studied 225 consecutive patients who underwent dynamic rest/stress rubidium-82 chloride (82Rb) PET imaging. Dynamic image series consisting of 30 frames were reconstructed with frame durations ranging from 5 to 80 seconds. An automated algorithm localized the RV and LV blood pools in space and time and then registered each frame to a tissue reference image volume using normalized gradient fields with a modification of a signed distance function. The computed shifts and their global and regional flow estimates were compared to those of reference shifts that were assessed by three physician readers. RESULTS: The automated motion-correction shifts were within 5 mm of the manual motion-correction shifts across the entire sequence. The automated and manual motion-correction global MBF values had excellent linear agreement (R = 0.99, y = 0.97x + 0.06). Uncorrected flows outside of the limits of agreement with the manual motion-corrected flows were brought into agreement in 90% of the cases for global MBF and in 87% of the cases for global MFR. The limits of agreement for stress MBF were also reduced twofold globally and by fourfold in the RCA territory. CONCLUSIONS: An image-based, automated motion-correction algorithm for dynamic PET across the entire dynamic sequence using normalized gradient fields matched the results of manual motion correction in reducing bias and variance in MBF and MFR, particularly in the RCA territory.

Identification and Quantification of Cardiovascular Structures From CCTA: An End-to-End, Rapid, Pixel-Wise, Deep-Learning Method.

OBJECTIVES: This study designed and evaluated an end-to-end deep learning solution for cardiac segmentation and quantification. BACKGROUND: Segmentation of cardiac structures from coronary computed tomography angiography (CCTA) images is laborious. We designed an end-to-end deep-learning solution. METHODS: Scans were obtained from multicenter registries of 166 patients who underwent clinically indicated CCTA. Left ventricular volume (LVV) and right ventricular volume (RVV), left atrial volume (LAV) and right atrial volume (RAV), and left ventricular myocardial mass (LVM) were manually annotated as ground truth. A U-Net-inspired, deep-learning model was trained, validated, and tested in a 70:20:10 split. RESULTS: Mean age was 61.1 ± 8.4 years, and 49% were women. A combined overall median Dice score of 0.9246 (interquartile range: 0.8870 to 0.9475) was achieved. The median Dice scores for LVV, RVV, LAV, RAV, and LVM were 0.938 (interquartile range: 0.887 to 0.958), 0.927 (interquartile range: 0.916 to 0.946), 0.934 (interquartile range: 0.899 to 0.950), 0.915 (interquartile range: 0.890 to 0.920), and 0.920 (interquartile range: 0.811 to 0.944), respectively. Model prediction correlated and agreed well with manual annotation for LVV (r = 0.98), RVV (r = 0.97), LAV (r = 0.78), RAV (r = 0.97), and LVM (r = 0.94) (p < 0.05 for all). Mean difference and limits of agreement for LVV, RVV, LAV, RAV, and LVM were 1.20 ml (95% CI: -7.12 to 9.51), -0.78 ml (95% CI: -10.08 to 8.52), -3.75 ml (95% CI: -21.53 to 14.03), 0.97 ml (95% CI: -6.14 to 8.09), and 6.41 g (95% CI: -8.71 to 21.52), respectively. CONCLUSIONS: A deep-learning model rapidly segmented and quantified cardiac structures. This was done with high accuracy on a pixel level, with good agreement with manual annotation, facilitating its expansion into areas of research and clinical import.

The utility of 82Rb PET for myocardial viability assessment: Comparison with perfusion-metabolism 82Rb-18F-FDG PET.

BACKGROUND: 82Rb kinetics may distinguish scar from viable but dysfunctional (hibernating) myocardium. We sought to define the relationship between 82Rb kinetics and myocardial viability compared with conventional 82Rb and 18F-fluorodeoxyglucose (FDG) perfusion-metabolism PET imaging. METHODS: Consecutive patients (N = 120) referred for evaluation of myocardial viability prior to revascularization and normal volunteers (N = 37) were reviewed. Dynamic 82Rb 3D PET data were acquired at rest. 18F-FDG 3D PET data were acquired after metabolic preparation using a standardized hyperinsulinemic-euglycemic clamp. 82Rb kinetic parameters K1, k2, and partition coefficient (KP) were estimated by compartmental modeling RESULTS: Segmental 82Rb k2 and KP differed significantly between scarred and hibernating segments identified by Rb-FDG perfusion-metabolism (k2, 0.42 ± 0.25 vs. 0.22 ± 0.09 min-1; P < .0001; KP, 1.33 ± 0.62 vs. 2.25 ± 0.98 ml/g; P < .0001). As compared to Rb-FDG analysis, segmental Rb KP had a c-index, sensitivity and specificity of 0.809, 76% and 84%, respectively, for distinguishing hibernating and scarred segments. Segmental k2 performed similarly, but with lower specificity (75%, P < .001) CONCLUSIONS: In this pilot study, 82Rb kinetic parameters k2 and KP, which are readily estimated using a compartmental model commonly used for myocardial blood flow, reliably differentiated hibernating myocardium and scar. Further study is necessary to evaluate their clinical utility for predicting benefit after revascularization.

Blood pool and tissue phase patient motion effects on 82rubidium PET myocardial blood flow quantification.

BACKGROUND: Patient motion can lead to misalignment of left ventricular volumes of interest and subsequently inaccurate quantification of myocardial blood flow (MBF) and flow reserve (MFR) from dynamic PET myocardial perfusion images. We aimed to identify the prevalence of patient motion in both blood and tissue phases and analyze the effects of this motion on MBF and MFR estimates. METHODS: We selected 225 consecutive patients that underwent dynamic stress/rest rubidium-82 chloride (82Rb) PET imaging. Dynamic image series were iteratively reconstructed with 5- to 10-second frame durations over the first 2 minutes for the blood phase and 10 to 80 seconds for the tissue phase. Motion shifts were assessed by 3 physician readers from the dynamic series and analyzed for frequency, magnitude, time, and direction of motion. The effects of this motion isolated in time, direction, and magnitude on global and regional MBF and MFR estimates were evaluated. Flow estimates derived from the motion corrected images were used as the error references. RESULTS: Mild to moderate motion (5-15 mm) was most prominent in the blood phase in 63% and 44% of the stress and rest studies, respectively. This motion was observed with frequencies of 75% in the septal and inferior directions for stress and 44% in the septal direction for rest. Images with blood phase isolated motion had mean global MBF and MFR errors of 2%-5%. Isolating blood phase motion in the inferior direction resulted in mean MBF and MFR errors of 29%-44% in the RCA territory. Flow errors due to tissue phase isolated motion were within 1%. CONCLUSIONS: Patient motion was most prevalent in the blood phase and MBF and MFR errors increased most substantially with motion in the inferior direction. Motion correction focused on these motions is needed to reduce MBF and MFR errors.