Predictive Accuracy Index Research Articles

Optimizing Kernel Density Estimation Bandwidth for Road Traffic Accident Hazard Identification: A Case Study of the City of London

Road traffic accidents pose significant challenges to sustainable urban safety and intelligent transportation management. The effective hazard identification of crash hotspots is crucial in implementing targeted safety measures. A severity-weighted system was adopted to quantify crash hazard levels. Using 1059 valid crash records of the City of London, the spatial correlations of crash points were first examined via average nearest neighbor analysis. Then, the optimal KDE bandwidth was determined via ArcGIS’s automatic extraction method, multi-distance spatial cluster analysis, and incremental spatial autocorrelation (ISA) analysis. The predictive accuracy index (PAI) was used to evaluate the accuracy of KDE results at various bandwidths. The results revealed a clustered spatial distribution of crash points. The optimized KDE bandwidth obtained via ISA analysis was 134 m, and the yielded PAI was 4.381, indicating better predictive accuracies and balanced hotspot distributions and reflecting both local concentrations and the overall continuity of crash hazard hotspots. Applying this bandwidth to the validation data allowed the successful identification of most high-risk areas and potential crash hazard hotspots attributed to traffic environmental factors; this method exhibits reliability, accuracy, and robustness over medium to long time scales. This workflow can serve as an analytical template for assisting planners in improving the identification accuracy of hazard hotspots, thereby reducing crash occurrences, actively promoting sustainable traffic safety development, and providing valuable insights for targeted crash prevention and intelligent traffic safety management in urban areas.

Proteomic profiling improves prognostic risk stratification of the Sarculator nomogram in soft tissue sarcomas of the extremities and trunk wall.

High-risk soft tissue sarcomas of the extremities and trunk wall (eSTS), as defined by the Sarculator nomogram, are more likely to benefit from (neo)adjuvant anthracycline-based therapy compared to low/intermediate-risk patients. The biology underpinning these differential treatment outcomes remain unknown. We analysed proteomic profiles and clinical outcomes of 123 eSTS patients. A Cox model for overall survival including the Sarculator was fitted to individual data to define four risk groups. A DNA replication protein signature-Sarcoma Proteomic Module 6 (SPM6) was evaluated for association with clinicopathological factors and risk groups. SPM6 was added as a covariate together with Sarculator in a multivariable Cox model to assess improvement in prognostic risk stratification. DNA replication and cell cycle proteins were upregulated in high-risk versus very low-risk patients. Evaluation of the functional effects of CRISPR-Cas9 gene knockdown of proteins enriched in high-risk patients using the cancer cell line encyclopaedia database identified candidate drug targets. SPM6 was significantly associated with tumour malignancy grade (p = 1.6e-06), histology (p = 1.4e-05) and risk groups (p = 2.6e-06). Cox model analysis showed that SPM6 substantially contributed to a better calibration of the Sarculator nomogram (Index of Prediction Accuracy = 0.109 for Sarculator alone versus 0.165 for Sarculator + SPM6). Risk stratification of patient with STS is defined by distinct biological pathways across a range of cancer hallmarks. Incorporation of SPM6 protein signature improves prognostic risk stratification of the Sarculator nomogram. This study highlights the utility of integrating protein signatures for the development of next-generation nomograms.

Graph-Based Hotspot Detection of Socio-Economic Data Using Rough-Set

The term hotspot refers to a location or an area where the occurrence of a particular phenomenon, event, or activity is significantly higher than in the surrounding areas. The existing statistical methods need help working well on discrete data. Also, it can identify a false hotspot. This paper proposes a novel graph-based hotspot detection using a rough set (GBHSDRS) for detecting the hotspots. This algorithm works well with discrete spatial vector data. Furthermore, it removes the false hotspot by finding the statistical significance of the identified hotspots. A rough set theory is applied to the graph of the spatial polygon data, and the nodes are divided into lower, boundary, and negative regions. Therefore, the candidate hotspot belongs to the lower region of the set, and the boundary value analysis will ensure the identification of the hotspots if the hotspot is present in the dataset. The p-value is used to find the statistical significance of the hotspots. The algorithm is tested on the socioeconomic data of Uttar Pradesh (UP) from 1991 on medical facilities. The average gain in density and Hotspot Prediction Accuracy Index (HAPI) of the detected hotspots is 26.54% and 23.41%, respectively. An average reduction in runtime is 27.73%, acquired compared to all other methods on the socioeconomic data.

Predicting the risks of kidney failure and death in adults with moderate to severe chronic kidney disease: multinational, longitudinal, population based, cohort study

ObjectiveTo train and test a super learner strategy for risk prediction of kidney failure and mortality in people with incident moderate to severe chronic kidney disease (stage G3b to G4).DesignMultinational,...

Limited incremental predictive value of the frailty index and other vulnerability measures from routine care data for mortality risk prediction in older patients with COVID-19 in primary care

BackgroundDuring the COVID-19 pandemic, older patients in primary care were triaged based on their frailty or assumed vulnerability for poor outcomes, while evidence on the prognostic value of vulnerability measures in COVID-19 patients in primary care was lacking. Still, knowledge on the role of vulnerability is pivotal in understanding the resilience of older people during acute illness, and hence important for future pandemic preparedness. Therefore, we assessed the predictive value of different routine care-based vulnerability measures in addition to age and sex for 28-day mortality in an older primary care population of patients with COVID-19.MethodsFrom primary care medical records using three routinely collected Dutch primary care databases, we included all patients aged 70 years or older with a COVID-19 diagnosis registration in 2020 and 2021. All-cause mortality was predicted using logistic regression based on age and sex only (basic model), and separately adding six vulnerability measures: renal function, cognitive impairment, number of chronic drugs, Charlson Comorbidity Index, Chronic Comorbidity Score, and a Frailty Index. Predictive performance of the basic model and the six vulnerability models was compared in terms of area under the receiver operator characteristic curve (AUC), index of prediction accuracy and the distribution of predicted risks.ResultsOf the 4,065 included patients, 9% died within 28 days after COVID-19 diagnosis. Predicted mortality risk ranged between 7–26% for the basic model including age and sex, changing to 4–41% by addition of comorbidity-based vulnerability measures (Charlson Comorbidity Index, Chronic Comorbidity Score), more reflecting impaired organ functioning. Similarly, the AUC of the basic model slightly increased from 0.69 (95%CI 0.66 – 0.72) to 0.74 (95%CI 0.71 – 0.76) by addition of either of these comorbidity scores. Addition of a Frailty Index, renal function, the number of chronic drugs or cognitive impairment yielded no substantial change in predictions.ConclusionIn our dataset of older COVID-19 patients in primary care, the 28-day mortality fraction was substantial at 9%. Six different vulnerability measures had little incremental predictive value in addition to age and sex in predicting short-term mortality.

Assessing Crime History as a Predictor: Exploring Hotspots of Violent and Property Crime in Malmö, Sweden

Objectives: Assessing the predictive accuracy of using prior crime, place attributes, ambient population, community structural, and social characteristics, in isolation and combined when forecasting different violent and property crimes. Method: Using multilevel negative binomial regression, crime is forecasted into the subsequent year, in 50-m grid-cells. Incidence rate ratio (IRR), Prediction Accuracy Index (PAI), and Prediction Efficacy Index (PEI*) are interpreted for all combined crime generators and community characteristics. This study is partially a test of a crude version of the Risk Terrain Modeling technique. Results: Where crime has been in the past, the risk for future crime is higher. Where characteristics conducive to crime congregate, the risk for crime is higher. Community structural characteristics and ambient population are important for some crime types. Combining variables increases the accuracy for most crime types, looking at the IRR. Taking the geographical area into account, crime history in combination with both place- and neighborhood characteristics reaches similar accuracy as crime history alone for most crime types and most hotspot cutoffs. Conclusions: Crime history, place-, and neighborhood-level attributes are all important when trying to accurately forecast crime, long-term at the micro-place. Only counting past crimes, however, still does a really good job.

The diagnostic value of ultrasonography in evaluation of the intraneural vascular anomalies of peripheral nerves.

The imaging diagnosis of intraneural vascular anomalies of peripheral nerves mostly depended on magnetic resonance imaging (MRI), whereas high-frequency ultrasonography evaluation of intraneural vascular anomalies has been seldom done. To evaluate the diagnostic value of ultrasonography in the diagnosis of intraneural vascular anomalies of peripheral nerves. A total of 69 consecutive patients seen at Shandong Provincial Hospital Affiliated to Shandong First Medical University between February 2013 and June 2022, each with a clinical suspicion of intraneural vascular anomaly, were included. The ultrasonographic images of intraneural vascular anomalies of peripheral nerves were analyzed and the ultrasonographic features were summarized. These data were compared with MRI, which served as the gold standard for the diagnosis of intraneural vascular anomalies. The kappa statistic was adopted to determine the level of agreement. The sensitivity, specificity, positive and negative predictive values, accuracy, and Youden index of high-frequency ultrasonography as a diagnostic tool were assessed. Ultrasonography findings were positive in 20 of 69 patients with a clinical suspicion of intraneural vascular anomaly. The diagnosis was confirmed by MRI in 21 patients. There was one false-positive result and two false-negative results by ultrasonography. The κ value was 0.896. The sensitivity, specificity, positive and negative predictive values, accuracy, and Youden index of ultrasonography were 90.5%, 97.9%, 95%, 95.9%, 95.7%, and 0.884, respectively. Ultrasonography could be an accurate, reliable, and convenient imaging tool for the diagnosis of intraneural vascular anomalies of peripheral nerves.

Spatial Analysis of Urban Road Traffic Accidents Using GIS

Road traffic accidents are on the rise in urban areas globally, including Sri Lanka, leading to significant human and property losses. The Colombo Municipal Council (CMC) area in Sri Lanka witnesses a particularly high number of these accidents. This study aimed to analyze urban traffic accidents using GIS within complex urban networks. It employed a two-step approach, starting with a hot spot analysis based on accident times and types (fatal, non-grievous, and damage-only), using Kernel Density Estimation (KDE) and Nearest Neighbor Hierarchy (NNH) methods. Additionally, it created separate severity maps and compares the results visually, with KDE offering a comprehensive overview and NNH pinpointing high-accident areas. The KDE method was expected to benefit long-term traffic congestion reduction and safety enhancement. NNH's severity map provides immediate insights for road safety specialists. A mathematical approach involved calculating the Prediction Accuracy Index (PAI) for each method, revealing a notably higher value for NNH, making it a more suitable choice. In conclusion, this study recommended the Nearest Neighbor Hierarchy (NNH) method as superior for road safety analysis, offering important guidance for urban planning and accident prevention efforts

External Validation of the OAC3-PAD Bleeding Score in a Nationwide Population of Patients Undergoing Invasive Treatment for Peripheral Arterial Disease

The OAC3-PAD score was developed to predict bleeding risk in patients with lower extremity peripheral arterial disease (PAD), but its performance in concomitant international cohorts is largely unknown. This study aimed to validate the OAC3-PAD score in an unselected nationwide population of patients undergoing invasive treatment for symptomatic PAD. This was a nationwide cohort study including all patients who underwent a first revascularisation procedure or major amputation for symptomatic PAD in Denmark from 2000 - 2021. The study population was stratified based on OAC3-PAD score, and the one year risk of major bleeding was assessed, accounting for the competing risk of death. The score performance was evaluated using calibration plots, C statistic, Brier score, and the index of prediction accuracy (IPA). A total of 52 016 patients were included (mean age 71 years, 43.8% female). The one year risk of major bleeding increased with higher OAC3-PAD score, ranging from 1.6% (95% confidence interval [CI] 1.4 - 1.8%) to 2.3% (95% CI 2.0 - 2.5%), 3.5% (95% CI 3.2 - 3.8%), and 5.2% (95% CI 4.8 - 5.6%) for patients with low, low moderate, moderate high, and high score, respectively. Using patients with low risk as reference, the OAC3-PAD score effectively categorised patients, demonstrating statistically significant differences in bleeding risk acrossstrata. However, the score showed modest discriminative performance, with a C statistic of 65% (95%CI 63 - 66%) and a Brier score of 2.6% (95% CI 2.5 - 2.7%). Nevertheless, it performed significantly better than the null model, as indicated by an IPA of 3.1%. Among patients who underwent invasive treatment for symptomatic PAD in routine care, the OAC3-PAD score was associated with greater risk of major bleeding with increasing score level. However, its discriminatory ability was modest, and the clinical utility remains to be determined.

Investigating Causes of Model Instability: Properties of the Prediction Accuracy Index

The Prediction Accuracy Index (PAI) monitors stability, defined as whether the predictive power of a model has deteriorated due to a change in the distribution of the explanatory variables since its development. This paper shows how the PAI is related to the Mahalanobis distance, an established statistic for examining high leverage observations in data. This relationship is used to explore properties of the PAI, including tools for how the PAI can be decomposed into effects due to (a) individual observations/cases, (b) individual variables, and (c) shifts in the mean of variables. Not only are these tools useful for practitioners to help determine why models fail stability, but they also have implications for auditors and regulators. In particular, reasons why models containing econometric variables may fail stability are explored and suggestions to improve model development are discussed.

Development of a non-targeted approach using three handheld spectrometers combined with ensemble classifiers for authentication of bovine milk

In this study, three Vis/NIR handheld spectrometers coupled with four machine learning methods were developed for the analysis of the authentic bovine milk samples and adulterated samples with hydrogen peroxide and sodium hypochlorite. On this matter, ensemble bagged tree (EBT), random under sampling-boosted ensemble (RUS-BE), random subspace discriminant ensemble (RSDE), and random subspace ensemble k-nearest neighbor (RSE-kNN) were proposed and compared. The experiments were performed in two different scenarios. In the first scenario, pure milk and adulterated samples with one adulterant were analyzed separately with three handheld spectrometers. By comparing the results of ensemble classifiers for three spectrometers, RSE-KNN represented the best results for the accuracy and Youden index (96%, and 93% respectively). Interestingly, the results of RSE-kNN were much better than kNN as base learner. In the second scenario, the mixtures of two adulterants were analyzed with three handheld spectrometers. Among them, one presented the best results, with the predictive accuracy and Youden index of 100% for EBT, RSDE and RSE-kNN.

Validade relativa e reprodutibilidade de indicadores da OMS para avaliação da alimentação de crianças menores de dois anos

The study aimed to evaluate relative validity and reproducibility of seven WHO indicators of dietary practices in children aged 6-23.9 months. Data from probabilistic sample of children who used primary healthcare services in Rio de Janeiro, Brazil were collected using a 24h dietary recall (24HR) and a closed questionnaire (Q1) on feeding in the day before the study. The last one was reapplied (Q2) around 16 days later. Validity was assessed by comparing the prevalence rates estimated by 24HR and Q1 and calculating the positive (PPV) and negative (NPV) predictive values, sensitivity (Se), specificity (Sp), and accuracy index (AI) for the resulting indicators. For reproducibility, estimated prevalence rates based on Q1 and Q2 were compared and the kappa index and prevalence-adjusted bias-adjusted kappa were estimated. Of the seven estimated indicators, the prevalence of two was overestimated (Continued breastfeeding: 50.0% vs 40.0%; Sweet beverage consumption: 65.1% vs 52.7%) and the prevalence of one was underestimated (Zero vegetable or fruit consumption: 6.5% vs 18.1%). For most indicators, Se and PPV were higher than Sp and NPV. The prevalence rates determined with Q1 and Q2 were similar for 6 indicators. More than half showed good, very good or excellent agreement.

Predictive Accuracy Index in evaluating the dataset shift (case study)

A dataset shift takes place in a situation, where the joint distribution of inputs and outputs differs between the stages of training, testing, and using predictive models. If the distribution of current data for the implemented forecasting model changed significantly compared to the distribution of data used to develop it, them it could lead to its incorrect operation. The aim of the study was to compare the properties of two indicators, the Univariate Predictive Accuracy Index (UPAI) and Multivariate Predictive Accuracy Index (MPAI). The research procedure was carried out in 2 scenarios. The first involved a comparison of UPAI and MPAI for the distributions of categorical variables, while the second for distributions of continuous variables. The obtained MPAI values for both scenarios were summarized and compared with the UPAI and PSI values calculated and published in our previous article [1]. The results of the experiment proved that basing the decision on the need to calibrate the model or build a completely new model on the basis of MPAI, a measure that takes into account the multivariate distribution of variables, is superior to one-dimensional measures.

A Short-Term Power Load Forecasting Method of Based on the CEEMDAN-MVO-GRU

Given that the power load data are stochastic and it is difficult to obtain accurate forecasting results by a single algorithm. In this study, a combined forecasting method for short-term power load was proposed based on the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN), Multiverse optimization algorithm (MVO), and the Gated Recurrent Unit (GRU) based on Rectified Adam (RAdam) optimizer. Firstly, the model uses the CEEMDAN algorithm to decompose the original electric load data into subsequences of different frequencies, and the dominant factors are extracted from the subsequences. Then, a GRU network based on the RAdam optimizer was built to perform the forecasting of the subsequences using the existing subsequences data and the associated influencing factors as the data set. Meanwhile, the parameters of the GRU network were optimized with the MVO optimization algorithm for the prediction problems of different subsequences. Finally, the prediction results of each subsequence were superimposed to obtain the final prediction results. The proposed combined prediction method was implemented in a case study of a substation in Weinan, China, and the prediction accuracy was compared with the traditional prediction method. The prediction accuracy index shows that the Root Mean Square Error of the prediction results of the proposed model is 80.18% lower than that of the traditional method, and the prediction accuracy error is controlled within 2%, indicating that the proposed model is better than the traditional method. This will have a favorable impact on the safe and stable operation of the power grid.

Preoperative Neutrophil-to-Lymphocyte Ratio and a New Inflammatory Biomarkers Prognostic Index for Primary Retroperitoneal Sarcomas: Retrospective Monocentric Study.

Inflammatory biomarkers and neutrophil-to-lymphocyte ratio (NLR) are associated with prognosis in several tumors. Data on sarcomas are limited, and insufficient on retroperitoneal sarcoma (RPS). Patients with primary RPS operated between 2002 and 2016 were included. Hemoglobin, monocytes, NLR, platelet-to-lymphocyte ratio (PLR) were retrieved and analyzed both individually and combined into a prognostic index (IBPI). Correlation with clinicopathologic variables was studied, as well as postoperative morbidity according to NLR and IBPI risk categories. The association between overall survival (OS) and biomarkers and, in addition, the 7-year Sarculator-predicted OS probability (pOS) was analyzed using univariable and multivariable Cox models. 423/463 patients had complete data. The median follow-up was 84 months. The median NLR was 3.3 (IQR, 2.4-4.7), with significant variation across histologies. NLR was the only biomarker that independently predicted OS (HR, 1.2; 95% CI, 1.03-1.40; P = 0.02). The IBPI showed good discrimination for subgroups at different OS (log-rank test P < 0.0001). The Cox model for pOS alone showed a 7-year index of prediction accuracy of 26.9, which increased to 29.5 when IBPI was added to pOS as a complementary prognostic tool. IBPI was also associated with the risk of serious infectious postoperative complications (P = 0.0094; noninfectious complications, P = 0.6463). NLR was an independent prognostic factor for OS in RPS. When combined into a prognostic index with hemoglobin, monocytes, and PLR, it serves as a readily available prognostic tool addressing tumor-related inflammation and helps in classifying RPS risk in addition to the Sarculator nomogram.

Statistical Approach of Identifying Crime Hotspots for GIS Mapping in Sri Lanka

Currently, crimes are well organized and difficult to investigate, and hence, the situation is more complex due to technological advancements. Therefore, there is a timely need to plan strategies to reduce crime using historical crime data. The objectives of the study are extraction of crime patterns using trend analysis, identifying the relationship between crime and geographic environment, and recognizing the district where crime is most prevalent. Five types of crimes that have a high impact on society from 2010 to 2019 in Sri Lanka were identified using the high mean cluster. There were House Breaking (HB), Hurt by Knife (HK), Robbery (RB), Rape (RP), and Cheating (CH). The result of Shapiro-Wilk test for normality wasn’t normal therefore Kruskal-Wallis multiple comparisons were used. There was a significant difference between the mean of HB, HK, RB, RP, and CH. Using Spearman’s rank correlation coefficient test, a very strong positive correlation was found between RP and HB (0.9472), RB and HB (0.9003) and HK and HB (0.9277). The type of crimes scattered throughout Sri Lanka was analyzed using Geographic Information System (GIS). Thematic maps were generated to identify hotspot areas. The risk of all five types of crime appeared to be high in Kurunegala and Gampaha districts. Prediction Accuracy Index (PAI) values were calculated to compare the predictive accuracy of crime types and the highest PAI value (3.98) was for RB crime. According to the findings of this study, a new security strategy can be developed to eradicate these trends from society.

ACS-NSQIP – Surgical risk calculator accurately predicts outcomes of laparotomy in a prospective study at a tertiary hospital in Tanzania

Introduction: The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) surgical risk calculator is excellent in predicting postoperative complications. Validation in resource-limited settings and applicability in gastrointestinal surgery is still unclear. Objective: The study aimed to determine the accuracy of the ACS-NSQIP surgical risk calculator in predicting the 30days postoperative adverse outcomes among patients who underwent laparotomy. Methods: A Single hospital-based, prospective cohort study was done at Muhimbili National Hospital in Tanzania from April 2021 to December 2021 recruiting patients aged 18 and above who underwent primary laparotomy. SR calculator variables were obtained from patients and entered manually to categorize patients into low and high-risk. Patients were followed up for thirty days for the outcome. The predicted risk was compared with actual occurrence to obtain personal risk ratios. The c-statistics of &gt; 0.7, Brier of score &lt; 0.25 and Index of Prediction Accuracy score were used for discrimination, accuracy, and usefulness of the model respectively. Results: ACS -NSQIP SR-Calculator discriminated well the risks of cardiac complications, re-laparotomy, Anastomotic leak, and death (c-statistic &gt; 0.7) Poor discrimination was observed for the length of hospital stay (c-statistics 0.518). However, SR-calculator shows high calibration potential for all complications with a Brier score &lt; 0.25 (0.002 - 0.144) and an IPA score ranging from 0.225 - 0.969. Conclusion: The ACS -NSQIP SR-calculator accurately predicted postoperative outcomes for patients requiring laparotomy. SR-Calculator is a reliable tool for preoperative shared decision-making and counseling. The model should be adopted to strengthen the healthcare system in a low-income country. What is already known about this subject?  The ACS NSQIP surgical risk calculator is accurate in predicting the outcome of the Surgical procedure.  ACS NSQIP surgical risk calculator was validated in the USA  The SRC is associated with surgical care improvement whenever applied. What are the new findings?  The SRC accurately predicted the 30-day outcomes associated with primary laparotomy except for the length of hospital stay, discharge for other services, and veno-thromboembolic events.  The risk ratio (RR) provides a preliminary estimation of specific complications for a single patient.  Index of prediction Accuracy (IPA) reliably summarizes the discrimination and calibration of the ACS-NSQIP-SR-Calculator.  The SRC can be validated in a resource-limited setting and can be a source of surgical care improvement. How might these results affect future research or surgical practice?  The results of this study will provide a comparison between our institution and other international situations where a calculator had been tested and surgical service improvement locally. Our findings will provide insights into our loco-regional surgical institutions in the low-income country to conduct cross-specialties and institutional research toward the adoption of ACS-NSQIP-SRC into National surgical improvement programs. Informative to the surgical community on the performance of the SR-calculator outside the USA.

Adaptive Kernel Density Estimation for Traffic Accidents Based on Improved Bandwidth Research on Black Spot Identification Model

At present, the total length of accident blackspot accounts for 0.25% of the total length of the road network, while the total number of accidents that occurred at accident black spots accounts for 25% of the total number of accidents on the road network. This paper describes a traffic accident black spot recognition model based on the adaptive kernel density estimation method combined with the road risk index. Using the traffic accident data of national and provincial trunk lines in Shanghai and ArcGIS software, the recognition results of black spots were compared with the recognition results of the accident frequency method and the kernel density estimation method, and the clustering degree of recognition results of adaptive kernel density estimation method were analyzed. The results show that: the accident prediction accuracy index values of the accident frequency method, kernel density estimation method, and traffic accident black spot recognition model were 14.39, 16.36, and 18.25, respectively, and the lengths of the traffic accident black spot sections were 184.68, 162.45, and 145.57, respectively, which means that the accident black spot section determined by the accident black spot recognition model was the shortest and the number of traffic accidents identified was the largest. Considering the safety improvement budget of 20% of the road length, the adaptive kernel density estimation method could identify about 69% of the traffic accidents, which was 1.13 times and 1.27 times that of the kernel density estimation method and the accident frequency method, respectively.

Evaluation of Risk Scores as Predictive Tools for Stroke in Patients with Retinal Artery Occlusion: A Danish Nationwide Cohort Study.

Purpose We investigated the 1-year risk of stroke in patients with retinal artery occlusion and evaluated the predictive and discriminating abilities of contemporary risk stratification models for embolic stroke. Methods This register-based cohort study included 7,906 patients with retinal artery occlusion from Danish nationwide patient registries between 1995 and 2018. The study population was stratified according to the number of points obtained in the stroke risk scores: the CHA 2 DS 2 -VASc score and the ESSEN Stroke Risk score. The 1-year risk of stroke within strata was evaluated and compared using the cox proportional hazards model. Furthermore, the discrimination of the risk scores as predictive tools for stroke risk assessment was investigated using C-statistics, Brier score, and the index of prediction accuracy. Results The stroke event rate in patients with retinal artery occlusion increased as the score increased for both risk scores, ranging from 3.62 (95% confidence interval [CI]: 2.46-5.31) per 100 person-years to 13.25 (95% CI: 11.78--14.89) per 100-person-years for increasing levels of the CHA 2 DS 2 -VASc score and from 3.97 (95% CI: 2.97-5.32) per 100 person-years to 16.43 (95% CI: 14.01-19.27) per 100 person-years for increasing levels of the ESSEN Stroke Risk score. Using a risk score of 0 as a reference, the difference was statistically significant for retinal artery occlusion patients with a CHA 2 DS 2 -VASc score of 2 or above and for all levels of the ESSEN Stroke Risk score. The C-statistics for the risk scores was 61% (95% CI: 58%-63%) and 62% (95% CI: 59-64%) for the CHA 2 DS 2 -VASc score and ESSEN Stroke Risk score, respectively. Conclusion The results suggested that the use of the CHA 2 DS 2 -VASc score and the ESSEN Stroke Risk score was applicable for risk stratification of stroke in patients with retinal artery occlusion, but discrimination was poor due to low specificity.

Multi-Ethnic Study of Atherosclerosis: Relationship between Left Ventricular Shape at Cardiac MRI and 10-year Outcomes.

Background Left ventricular (LV) subclinical remodeling is associated with adverse outcomes and indicates mechanisms of disease development. Standard metrics such as LV mass and volumes may not capture the full range of remodeling. Purpose To quantify the relationship between LV three-dimensional shape at MRI and incident cardiovascular events over 10 years. Materials and Methods In this retrospective study, 5098 participants from the Multi-Ethnic Study of Atherosclerosis who were free of clinical cardiovascular disease underwent cardiac MRI from 2000 to 2002. LV shape models were automatically generated using a machine learning workflow. Event-specific remodeling signatures were computed using partial least squares regression, and random survival forests were used to determine which features were most associated with incident heart failure (HF), coronary heart disease (CHD), and cardiovascular disease (CVD) events over a 10-year follow-up period. The discrimination improvement of adding LV shape to traditional cardiovascular risk factors, coronary artery calcium scores, and N-terminal pro-brain natriuretic peptide levels was assessed using the index of prediction accuracy and time-dependent area under the receiver operating characteristic curve (AUC). Kaplan-Meier survival curves were used to illustrate the ability of remodeling signatures to predict the end points. Results Overall, 4618 participants had sufficient three-dimensional MRI information to generate patient-specific LV models (mean age, 60.6 years ± 9.9 [SD]; 2540 women). Among these participants, 147 had HF, 317 had CHD, and 455 had CVD events. The addition of LV remodeling signatures to traditional cardiovascular risk factors improved the mean AUC for 10-year survival prediction and achieved better performance than LV mass and volumes; HF (AUC, 0.83 ± 0.01 and 0.81 ± 0.01, respectively; P < .05), CHD (AUC, 0.77 ± 0.01 and 0.75 ± 0.01, respectively; P < .05), and CVD (AUC, 0.78 ± 0.0 and 0.76 ± 0.0, respectively; P < .05). Kaplan-Meier analysis demonstrated that participants with high-risk HF remodeling signatures had a 10-year survival rate of 56% compared with 95% for those with low-risk scores. Conclusion Left ventricular event-specific remodeling signatures were more predictive of heart failure, coronary heart disease, and cardiovascular disease events over 10 years than standard mass and volume measures and enable an automatic personalized medicine approach to tracking remodeling. © RSNA, 2022 Online supplemental material is available for this article.