Predictive Accuracy Index Research Articles

A comparison of survival models for prediction of eight-year revision risk following total knee and hip arthroplasty

BackgroundThere is increasing interest in the development and use of clinical prediction models, but a lack of evidence-supported guidance on the merits of different modelling approaches. This is especially true for time-to-event outcomes, where limited studies have compared the vast number of modelling approaches available. This study compares prediction accuracy and variable importance measures for four modelling approaches in prediction of time-to-revision surgery following total knee arthroplasty (TKA) and total hip arthroplasty (THA).MethodsThe study included 321,945 TKA and 151,113 THA procedures performed between 1 January 2003 and 31 December 2017. Accuracy of the Cox model, Weibull parametric model, flexible parametric model, and random survival forest were compared, with patient age, sex, comorbidities, and prosthesis characteristics considered as predictors. Prediction accuracy was assessed using the Index of Prediction Accuracy (IPA), c-index, and smoothed calibration curves. Variable importance rankings from the Cox model and random survival forest were also compared.ResultsOverall, the Cox and flexible parametric survival models performed best for prediction of both TKA (integrated IPA 0.056 (95% CI [0.054, 0.057]) compared to 0.054 (95% CI [0.053, 0.056]) for the Weibull parametric model), and THA revision. (0.029 95% CI [0.027, 0.030] compared to 0.027 (95% CI [0.025, 0.028]) for the random survival forest). The c-index showed broadly similar discrimination between all modelling approaches. Models were generally well calibrated, but random survival forest underfitted the predicted risk of TKA revision compared to regression approaches. The most important predictors of revision were similar in the Cox model and random survival forest for TKA (age, opioid use, and patella resurfacing) and THA (femoral cement, depression, and opioid use).ConclusionThe Cox and flexible parametric models had superior overall performance, although all approaches performed similarly. Notably, this study showed no benefit of a tuned random survival forest over regression models in this setting.

Modal decomposition-based hybrid model for stock index prediction

Stock index prediction is considered one of the most challenging issues in the financial sector owing to its noise, volatility, and instability. Traditional stock index prediction methods, such as statistical and machine learning methods, cannot achieve a high denoising effect, and also cannot mine enough data features from the stock data, resulting in a poor prediction performance. Deep learning has become an effective tool to predict non-stationary and nonlinear stock indices with strong learning ability. However, there is still room for prediction accuracy improvement if a single deep learning prediction model is replaced with a hybrid model. Therefore, this study proposes a novel deep learning hybrid model for stock index prediction named CEEMDAN-DAE-LSTM. In this hybrid model, the stock index is first decomposed using complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) into a series of intrinsic mode functions (IMFs) arranged from high to low frequency. Next, the deep autoencoder (DAE) is applied to remove redundant data and extract deep-level features. Then, high-level abstract features are separately fed into long short-term memory (LSTM) networks to predict the stock returns of the next trading day. Finally, the final predicted value is obtained by synthesizing the value of each component. Empirical research results on six stock indices representing both developed and emerging markets showed that our model is superior to other reference models in terms of prediction accuracy and stock index trends; furthermore, it has higher prediction performance for stock indices with greater volatility. In general, this model could be applicable to various stock markets with different degrees of development.

Efficacy of Plasmodium falciparum histidine-rich protein 2 ( Pfhrp 2) rapid diagnostic test (RDT) and microscopy in the detection of falciparum malaria among symptomatic patients in Akure, Nigeria.

Accurate diagnosis and prompt treatment are highly essential in the management of malaria, which is one of the deadliest infectious diseases worldwide, particularly in tropical and sub-tropical regions including Nigeria. This study was designed to evaluate the efficacy of malaria histidine-rich protein 2-based rapid diagnostic test (RDT) and microscopy in the diagnosis of falciparum malaria in Nigeria. This was a cross-sectional and hospital-based study. The standard method of microscopy was used as the gold standard. Giemsa stained thick and thin smears were prepared to count and detect malaria parasite species. Also, a malaria histidine-rich protein 2-based RDT was used to detect malaria parasites and diagnostic efficacy were determined through the measure of sensitivity, specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV), diagnostic accuracy and Youden Index (J). The result showed that out of the total 303 individuals examined, a total malaria prevalence of 67.0% and 68.0% were recorded for micr oscopy and RDT, respectively. Additionally, the sensitivity (95% C.I), specificity (95% C.I), PPV (95% C.I), and NPV (95% C.I) of RDT compared to microscopy were 97.54 (94.36-98.94), 92.00 (85.00-95.89), 96.12 (92.53-98.02), and 94.85 (88.50- 97.78), respectively. The diagnostic accuracy (95% C.I) and Youden Index (J) were 95.71 (92.77- 97.70) and 0.89, respectively. Conclusively, our study revealed that RDT continues to remain efficacious. Thus, while malaria diagnosis by microscopy which is the gold standard remains the major method of malaria detection, it should be complemented by rapid diagnostic test (RDT), particularly in high malaria endemic regions where mean parasite density of patients are usually high.

An integrated solution to identify pedestrian-vehicle accident prone locations: GIS-based multicriteria decision approach

Spatial distributions of pedestrian-vehicle accident-prone locations (APLs) according to GIS-based models differ. Also, which APLs are determined by conventional models are more critical or which model is more successful in determining APL is still a major concern. To bridge this gap, this paper presents an innovative GIS-based Multi-Criteria Decision Making (MCDM) approach to identify the most critical APLs and to rank APLs with the compromising results of four GIS-based models. The results of planar KDE, network-based KDE, Getis-Ord Gi*, and Local Moran’s I which are weighted with prediction accuracy index (PAI), were evaluated together with MCDM methods: traditional VIKOR and psychometric VIKOR. Results & Discussion: The 15 most critical APLs in the compromise solution were ranked for four time periods. Network-based KDE gave the best performance, while Local Moran’s I performed the worst. Sensitivity analysis showed that the Psychometric VIKOR provides acceptable stability in the rankings of the APLs. The innovative MCDM approaches allowed the results of several models to be evaluated together. Thus, more reliable APLs were identified. Local governments with limited budgets can determine which APLs should be considered to improve pedestrian safety with the recommended approach and can apply to any study area.

Multitask Learning Radiomics on Longitudinal Imaging to Predict Survival Outcomes following Risk-Adaptive Chemoradiation for Non-Small Cell Lung Cancer.

Simple SummaryPersonalized cancer treatment strategies, including risk-adaptive chemoradiation therapy based on medical imaging, seek to improve outcomes of patients with unresectable and locally advanced non-small cell lung cancer. Refining patient risk stratification relies on outcome prediction modeling based in part on information from different imaging modalities and imaging time points during and after treatment. Using prospectively collected longitudinal data from FDG-PET, CT, and perfusion SPECT images of patients enrolled on a clinical trial, our aim was to evaluate the utility of a multitask machine learning radiomics framework for survival outcome prediction. We found that multitask learning of FDG-PET radiomics on pretreatment and mid-treatment images achieved higher survival prediction concordance compared with single-task learning of other modalities and clinical benchmark models. Our multitask learning radiomics framework can be applied to other longitudinal imaging datasets, and, once validated, can strengthen clinical decision support for personalized and adaptive treatment courses.Medical imaging provides quantitative and spatial information to evaluate treatment response in the management of patients with non-small cell lung cancer (NSCLC). High throughput extraction of radiomic features on these images can potentially phenotype tumors non-invasively and support risk stratification based on survival outcome prediction. The prognostic value of radiomics from different imaging modalities and time points prior to and during chemoradiation therapy of NSCLC, relative to conventional imaging biomarker or delta radiomics models, remains uncharacterized. We investigated the utility of multitask learning of multi-time point radiomic features, as opposed to single-task learning, for improving survival outcome prediction relative to conventional clinical imaging feature model benchmarks. Survival outcomes were prospectively collected for 45 patients with unresectable NSCLC enrolled on the FLARE-RT phase II trial of risk-adaptive chemoradiation and optional consolidation PD-L1 checkpoint blockade (NCT02773238). FDG-PET, CT, and perfusion SPECT imaging pretreatment and week 3 mid-treatment was performed and 110 IBSI-compliant pyradiomics shape-/intensity-/texture-based features from the metabolic tumor volume were extracted. Outcome modeling consisted of a fused Laplacian sparse group LASSO with component-wise gradient boosting survival regression in a multitask learning framework. Testing performance under stratified 10-fold cross-validation was evaluated for multitask learning radiomics of different imaging modalities and time points. Multitask learning models were benchmarked against conventional clinical imaging and delta radiomics models and evaluated with the concordance index (c-index) and index of prediction accuracy (IPA). FDG-PET radiomics had higher prognostic value for overall survival in test folds (c-index 0.71 [0.67, 0.75]) than CT radiomics (c-index 0.64 [0.60, 0.71]) or perfusion SPECT radiomics (c-index 0.60 [0.57, 0.63]). Multitask learning of pre-/mid-treatment FDG-PET radiomics (c-index 0.71 [0.67, 0.75]) outperformed benchmark clinical imaging (c-index 0.65 [0.59, 0.71]) and FDG-PET delta radiomics (c-index 0.52 [0.48, 0.58]) models. Similarly, the IPA for multitask learning FDG-PET radiomics (30%) was higher than clinical imaging (26%) and delta radiomics (15%) models. Radiomics models performed consistently under different voxel resampling conditions. Multitask learning radiomics for outcome modeling provides a clinical decision support platform that leverages longitudinal imaging information. This framework can reveal the relative importance of different imaging modalities and time points when designing risk-adaptive cancer treatment strategies.

Kriging Model for Reliability Analysis of the Offshore Steel Trestle Subjected to Wave and Current Loads

Offshore steel trestles (OSTs) are exposed to severe marine environments with stochastic wave and current loads, making structural safety assessment challenging and difficult. Reliability analysis is a suitable way to consider both wave and current loading intensity uncertainties, but the implicit and complex limit state functions of the reliability analysis usually imply huge computational costs. This paper proposes an efficient reliability analysis framework for OST using the kriging model of optimal linear unbiased estimation. The surrogate model is built with stochastic waves, current parameters, and the corresponding load factors. The framework is then used to evaluate the reliability of an example OST subjected to wave and current loads at three limit states of OST, including first yield (FY), full plastic (FP), and collapse initiation (CI). Three different distributions are used for comparison of the results of failure probability and reliability index. The results and the computational cost by the proposed framework are compared with that from the Monte Carlo sampling (MCS) and Latin hypercube sampling (LHS) method. The influences of sample number on the prediction accuracy and reliability index are investigated. The influence of marine growth on the reliability analysis of the OST is discussed using MCS and the kriging model. The results show that the reliability analysis based on the kriging model can obtain the reliability index for the OST efficiently with less calculation time but similar results compared with MCS and LHS. With the increase of the number of samples, the prediction accuracy of the kriging model increases, and the corresponding failure probability fluctuates greatly at first and then tends to be stable. The reliability of the example OST is reduced with the increase of marine growth, regardless of the limit state.

Inaccurate Risk Assessment by the ACS NSQIP Risk Calculator in Aortic Surgery.

Objectives: The aim of this retrospective study was to assess the predictive performance of the American College of Surgeons (ACS) risk calculator for aortic aneurysm repair for the patient population of a Dutch tertiary referral hospital. Methods: This retrospective study included all patients who underwent elective endovascular or open aortic aneurysm repair at our institution between the years 2013 and 2019. Preoperative patient demographics and postoperative complication data were collected, and individual risk assessments were generated using five different current procedural terminology (CPT) codes. Receiver operating characteristic (ROC) curves, calibration plots, Brier scores, and Index of Prediction Accuracy (IPA) values were generated to evaluate the predictive performance of the ACS risk calculator in terms of discrimination and calibration. Results: Two hundred thirty-four patients who underwent elective endovascular or open aortic aneurysm repair were identified. Only five out of thirteen risk predictions were found to be sufficiently discriminative. Furthermore, the ACS risk calculator showed a structurally insufficient calibration. Most Brier scores were close to 0; however, comparison to a null model though IPA-scores showed the predictions generated by the ACS risk calculator to be inaccurate. Overall, the ACS risk calculator showed a consistent underestimation of the risk of complications. Conclusions: The ACS risk calculator proved to be inaccurate within the framework of endovascular and open aortic aneurysm repair in our medical center. To minimize the effects of patient selection and cultural differences, multicenter collaboration is necessary to assess the performance of the ACS risk calculator in aortic surgery.

Using Model Performance to Assess the Representativeness of Data for Model Development and Calibration in Financial Institutions

This paper proposes a methodology that utilises model performance as a metric to assess the representativeness of external or pooled data when it is used by banks in regulatory model development and calibration. There is currently no formal methodology to assess representativeness. The paper provides a review of existing regulatory literature on the requirements of assessing representativeness and emphasises that both qualitative and quantitative aspects need to be considered. We present a novel methodology and apply it to two case studies. We compared our methodology with the Multivariate Prediction Accuracy Index. The first case study investigates whether a pooled data source from Global Credit Data (GCD) is representative when considering the enrichment of internal data with pooled data in the development of a regulatory loss given default (LGD) model. The second case study differs from the first by illustrating which other countries in the pooled data set could be representative when enriching internal data during the development of a LGD model. Using these case studies as examples, our proposed methodology provides users with a generalised framework to identify subsets of the external data that are representative of their Country’s or bank’s data, making the results general and universally applicable.

Accounting for previous events to model and predict traffic accidents at the road segment level: A study in Valencia (Spain)

Predicting the occurrence of traffic accidents is essential for establishing preventive measures and reducing the impact of traffic accidents. In particular, it is fundamental to make predictions using fine spatio-temporal units. In this paper, the daily risk of traffic accident occurrence across the road network of Valencia (Spain) is modeled through logistic regression models. The spatio-temporal dependence between the observations is accounted for through the inclusion of lagged binary covariates representing the previous occurrence of a traffic accident within a spatio-temporal window centered at each combination of day and segment of the network. A temporal distance of 28 days and a fifth-order spatial distance are set as the limits of such dependence. Furthermore, the models include fixed effects in terms of several socio-demographic, network-related, and weather-related covariates. Temporal (month and day of the week) and spatial (borough-level) effects are also considered. The predictive quality of the models is examined through the Matthews correlation coefficient and the prediction accuracy index. The results indicate that the incorporation of spatio-temporal dependence improves the predictive ability of the models. However, while the inclusion of temporally-lagged covariates representing short-and mid-term temporal dependence yields more accurate predictions, the higher-order spatial lags barely alter model performance.

Considerations for Developing Predictive Spatial Models of Crime and New Methods for Measuring Their Accuracy

A literature review of the important trends in predictive crime modeling and the existing measures of accuracy was undertaken. It highlighted the need for a robust, comprehensive and independent evaluation and the need to include complementary measures for a more complete assessment. We develop a new measure called the penalized predictive accuracy index (PPAI), propose the use of the expected utility function to combine multiple measures and the use of the average logarithmic score, which measures accuracy differently than existing measures. The measures are illustrated using hypothetical examples. We illustrate how PPAI could identify the best model for a given problem, as well as how the expected utility measure can be used to combine different measures in a way that is the most appropriate for the problem at hand. It is important to develop measures that empower the practitioner with the ability to input the choices and preferences that are most appropriate for the problem at hand and to combine multiple measures. The measures proposed here go some way towards providing this ability. Further development along these lines is needed.

Land Cover Classification of Remote Sensing Image of Hubei Province by Using PSP Net

Objectives Pyramid scene parsing net (PSP Net), a kind of neural network with deep structure, extracts the features of remote sensing image better than the traditional model, e.g. artificial neural network, and supports vector machine. This algorithm can embed the contextual features of difficult scenes into the pixel prediction framework of the full convolutional neural network (FCN), fully understand the scene, realize accurate prediction of each pixel category, location and shape, and fuse local and global information together to propose an optimization strategy for moderate supervised loss. A deep learning algorithm based on PSP Net is proposed to achieve higher accuracy in image classification and effectively promote remote sensing image automation and intelligent interpretation. Methods Conventional architecture for fast feature embedding (CAFFE) is a deep learning framework with is expressive, fast and thought modular, and supports a variety of deep learning architectures for image classification and segmentation. PSP Net model under the CAFFE framework is used by modifying and optimizing the network to eliminate the overfitting effect. Using the remote sensing image of Hubei Province as the experimental data, the land with the 30 m resolution is studied with the aid of the analysis ability of the context scene of PSP Net. The Python program is used in the CAFFE depth learning framework for operation. The operation network obtains the global feature information through the core structure (pyramid pooling module) and completes the generation of the data set. In the experiment, the 507 standard partial images of 900×600 pixels in the landsat image of Hubei Province are used, and the sample sets suitable for depth learning are generated by pre-processing. Among these, 300 data are selected, including 240 training sets, 44 prediction sets, and 16 validation sets. The overall accuracy (OA) index is used as the prediction accuracy index of the preliminary evaluation model to evaluate the accuracy of the prediction model. Results PSP Net model under the deep learning framework of CAFFE is used to train the sample data. The learning rate of 10×10-10 is set, and the million times training model is selected to protect against overfitting of the data. Through the generalization of the model and the generalization and iteration of the samples, the land cover of the 3 phases of TM images about Hubei Province are classified, and the classification accuracy is 82.2%, 83.4% and 83.7%, respectively. Conclusions The results show that the land cover classification of remote sensing images can be realized quickly, effectively and accurately by PSP Net.

External Validation of the VOCAL‐Penn Cirrhosis Surgical Risk Score in 2 Large, Independent Health Systems

Cirrhosis poses an increased risk of postoperative mortality, yet it remains challenging to accurately risk stratify patients in clinical practice. The VOCAL‐Penn cirrhosis surgical risk score was recently developed and internally validated in the national Veterans Affairs health system; however, to date this score has not been evaluated in independent cohorts. The goal of this study was to compare the predictive performance of the VOCAL‐Penn to the Mayo risk, Model for End‐Stage Liver Disease (MELD), and MELD‐sodium (MELD‐Na) scores in 2 large health systems. We performed a retrospective cohort study of patients with cirrhosis undergoing surgical procedures of interest at the Beth Israel Deaconess Medical Center or University of Pennsylvania Health System from January 1, 2008, to October 1, 2015. The outcomes of interest were 30‐day and 90‐day postoperative mortality. Concordance statistics (C‐statistics), calibration curves, Brier scores, and the index of prediction accuracy (IPA) were compared for each predictive model. A total of 855 surgical procedures were identified. The VOCAL‐Penn score had the numerically highest C‐statistic for 90‐day postoperative mortality (eg, 0.82 versus 0.79 Mayo versus 0.78 MELD‐Na versus 0.79 MELD), although differences were not statistically significant. Calibrations were excellent for the VOCAL‐Penn, MELD, and MELD‐Na; however, the Mayo score consistently overestimated risk. The VOCAL‐Penn had the lowest Brier score and highest IPA at both time points, suggesting superior overall predictive model performance. In subgroup analyses of patients with higher MELD scores, the VOCAL‐Penn had significantly higher C‐statistics compared with the MELD and MELD‐Na. The VOCAL‐Penn score (www.vocalpennscore.com) has excellent discrimination and calibration for postoperative mortality among diverse patients with cirrhosis. Overall performance is superior to the Mayo, MELD, and MELD‐Na scores. In contrast to the MELD/MELD‐Na, the VOCAL‐Penn retains excellent discrimination among patients with higher MELD scores.

Postanesthetic Cold Sensibility Test as an Indicator for the Efficacy of Inferior Alveolar Nerve Block in Patients with Symptomatic Irreversible Pulpitis of Mandibular Molars.

Materials and Methods This study was conducted on the mandibular first molars of 54 patients (35 males and 19 females) with signs and symptoms of SIP. To anesthetize the affected molars, all patients received a single carpule of 2% lidocaine with 1 : 100000 epinephrine using a standardized inferior alveolar nerve block (IANB) technique. The cold test was conducted before beginning the endodontic procedures and after gaining lip numbness, and the results were reported as either positive or negative response. The root canal preparation (RCP) was then initiated and the patients' responses were documented (Gold standard test). True pulpal anesthetic failure was described as a pain perception during the access cavity and pulp tissue removal. True pulpal anesthesia was defined as no pain or discomfort during the access cavity and pulp extirpation. The qualitative variables frequencies and percentages of patients with true/false positive and negative responses were determined and then compared using the Chi-square test. The pain perception of male and female patients during the cold test and gold standard was compared using the Fisher exact test. The following diagnostic parameters were calculated using an online statistical calculator: sensitivity, specificity, predictive values, accuracy, and Youden index. In addition, a receiver operating characteristic curve (ROC) was constructed and the area under the curve (AUC) was calculated. Results The overall percentage of actual failure of pupal anesthesia was 57%. The sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and Youden index for the cold test were 0.87, 0.91, 0.93, 0.84, 0.89, and 0.78, respectively. There was no statistically significant difference between male and female patients regarding their responses to cold testing and the gold standard test (P > 0.05). Besides, the patients' reactions to the cold test were significantly matched with their reactions to the gold standard test (P < 0.05). The area under the ROC was mostly 0.9. Conclusion The cold test could be a valuable and accurate method for predicting the potential pupal anesthesia before beginning the endodontic treatment of mandibular molars with symptomatic irreversible pulpitis, particularly after obtaining postanesthetic soft tissue numbness.

Accident Hotspots in Southern Expressway of Sri Lanka: Interpolation Evaluation using GIS

In this study, the southern expressway, which is the first and lengthiest E class highway (126 km) in Sri Lanka, was analysed for roadside accident incidences. The primary objective of this paper is to identify the best-fit interpolation techniques for the hotspots' most distinctive causes of vehicular crashes. The accident details were collected from the Police Headquarters consisting of 966 accidents that took place during the period from 2015 to 2017. To identify accident hotspots, GIS-based interpolation techniques such as Ordinary Kriging, Kernel Density Estimation (KDE), Inverse Distance Weighting (IDW), and Nearest Neighbour Interpolation methods were used. The spatial interpolation outcome of the four methods was compared based on the standard Prediction Accuracy Index (PAI). The analysis was executed using QGIS and GeoDa. Results of PAI revealed that an IDW and KDE outperformed the other two interpolation methods. The left and right lanes of the expressway, spotted with 11 and 20 hotspots, respectively, indicate the right lane was 50% more prone to accidents than the left lane. Notably, nearly 5% of the entire road stretch is estimated as accident-prone spots in both lanes. Peak accidents were recorded during afternoon and evening hours, and buses were the most active vehicle type. Uncontrolled speeding was the primary reason for more than 50% of the accidents, while unsuccessful overtake accounted for more than 20% of the accidents on the highway. The road design modifications and warning sign placements at appropriate places may be recommended as countermeasures.

GIS-based crash hotspot identification: a comparison among mapping clusters and spatial analysis techniques

Knowing the locations of traffic crash hotspots can provide us with valuable insights into the root causes of crash occurrence over the area under study. This knowledge helps decision-makers to better assess the risk associated with road crashes and, as a result, help them to propose more effective countermeasures in order to reduce the annual crash rate. Nonetheless, identifying the areas with the highest potential of crash occurrence is a complicated task. In this regard and within this study, five various types of hotspot identification techniques, consisting of Average Nearest Neighbor, Getis-Ord Gi*, Global Moran’s I, kernel density estimation (KDE) and mean centre, were compared to each other, using three different performance measures, including Predictive Accuracy Index (PAI), Recapture Rate Index (RRI) and hit rate. According to the results, the most accurate model with the highest PAI values (PAI = 1.61 and 1.76), Moran’s I, had the third-highest reliability value (RRI = 1.003). On the other hand, while the Gi* method was the most precise and reliable technique with the highest RRI value (RRI = 1.121), it showed the second-lowest accuracy (PAI= 0.83 and 0.74). Overall, it seems that Moran’s I method is superior to other methods in locating hotspots, which is not only the most accurate technique but also precise enough to rely on.

A Novel Validated Recurrence Stratification System Based on 18F-FDG PET/CT Radiomics to Guide Surveillance After Resection of Pancreatic Cancer.

objectiveDespite the heterogeneous biology of pancreatic cancer, similar surveillance schemas have been used. Identifying the high recurrence risk population and conducting prompt intervention may improve prognosis and prolong overall survival.MethodsOne hundred fifty-six resectable pancreatic cancer patients who had undergone 18F-FDG PET/CT from January 2013 to December 2018 were retrospectively reviewed. The patients were categorized into a training cohort (n = 109) and a validation cohort (n = 47). LIFEx software was used to extract radiomic features from PET/CT. The risk stratification system was based on predictive factors for recurrence, and the index of prediction accuracy was used to reflect both the discrimination and calibration.ResultsOverall, seven risk factors comprising the rad-score and clinical variables that were significantly correlated with relapse were incorporated into the final risk stratification system. The 1-year recurrence-free survival differed significantly among the low-, intermediate-, and high-risk groups (85.5, 24.0, and 9.1%, respectively; p < 0.0001). The C-index of the risk stratification system in the development cohort was 0.890 (95% CI, 0.835–0.945).ConclusionThe 18F-FDG PET/CT-based radiomic features and clinicopathological factors demonstrated good performance in predicting recurrence after pancreatectomy in pancreatic cancer patients, providing a strong recommendation for an adequate adjuvant therapy course in all patients. The high-risk recurrence population should proceed with closer follow-up in a clinical setting.

An Adaptive Spatial Resolution Method Based on the ST-ResNet Model for Hourly Property Crime Prediction

Effective predictive policing can guide police patrols and deter crime. Hourly crime prediction is expected to save police time. The selection of spatial resolution is important due to its strong relationship with the accuracy of crime prediction. In this paper, we propose an adaptive spatial resolution method to select the best spatial resolution for hourly crime prediction. The ST-ResNet model is applied to predict crime risk, with historical crime data and weather data as predictive variables. A prediction accuracy index (PAI) is used to evaluate the accuracy of the results. Data on property crimes committed in Suzhou, a big city in China, were selected as the research data. The experiment results indicate that a 2.4 km spatial resolution leads to the best performance for crime prediction. The adaptive spatial resolution method can be used to guide police deployment.

A Study on the Effect of Stock Price Prediction on Marketing Performance

This study examined how marketing performance affects company stock values based on the stock price prediction method. A causal relationship exists between marketing performance and stock value. Many factors (both external and internal) affect stock value, therefore, improving the accuracy of stock price forecasts is vital. In the digital economy, customers are connected via social media, thus word of mouth advocating a company has an important positive impact on marketing performance which, in turn, enhances stock value, making it essential to predict stock prices accurately. In this study, the improvement in the stock price prediction rate is explained using deep learning methods. The accuracy of stock price prediction is crucial to determine the value of a stock as it reflects the company's value. Therefore, we considered representative methods of deep learning to predict marketing performance in the context of developing stock price forecasting accuracy. In addition, this study improves stock price forecasting using a review of the mean squared error as an index of prediction accuracy using deep learning methods.

Development of a multivariable model predicting mortality risk from comorbidities in an Italian cohort of 18,286 confirmed COVID-19 cases aged 40 years or older.

to develop a risk prediction model for 30-day mortality from COVID‑19 in an Italian cohort aged 40 years or older. a population-based retrospective cohort study on prospectively collected data was conducted. the cohort included all swab positive cases aged 40 years older (No. 18,286) among residents in the territory of the Milan's Agency for Health Protection (ATS-MI) up to 27.04.2020. Data on comorbidities were obtained from the ATS administrative database of chronic conditions. to predict 30-day mortality risk, a multivariable logistic regression model, including age, gender, and the selected conditions, was developed following the TRIPOD guidelines. Discrimination and calibration of the model were assessed. after age and gender, the most important predictors of 30-day mortality were diabetes, tumour in first-line treatment, chronic heart failure, and complicated diabetes. The bootstrap-validated c-index was 0.78, which suggests that this model is useful in predicting death after COVID-19 infection in swab positive cases. The model had good discrimination (Brier score 0.13) and was well calibrated (Index of prediction accuracy of 14.8%). a risk prediction model for 30-day mortality in a large COVID-19 cohort aged 40 years or older was developed. In a new epidemic wave, it would help to define groups at different risk and to identify high-risk subjects to target for specific prevention and therapeutic strategies.

Community security is the key to sustainable governance: Methods and functions of crime hotspot predictions

Forecasting is becoming increasingly important in corporate sustainability governance, as is government governance, and the prediction of police crime hotspots is related to human rights, so transparency is needed. There are many ways to predict hotspots of criminal activity in urban areas. Experts assume that if many crimes occur somewhere, even more, are likely to happen at subsequent times. Such predictions may rely on a state dependency model such as the Poisson distribution algorithm to formulate re-occurrence, its results can provide a visualized hotspot map with Q-GIS maps. Forecasting sets the threshold for re-occurrence and affects the distribution of the forecast. This paper studies the occurrence of criminal activity in urban areas, refers to the metrics set by the NIJ’s crime prediction contest and focuses on the presentation of the results by accumulating different historical data. It was determined that when the amount of cumulative data is greater, its prediction measures by the prediction accuracy index (PAI) insures that accuracy is improved, but the prediction efficiency index (PEI) that efficiency level is worse. Because threshold setting directly affects the performance of the forecast, it can be used differently. Here sets four different indicators, hit rate, useful rate, waste rate, and missing rate. It was determined that the hit rate, missing rate, the PAI value, and the PEI value are directly proportional to the threshold value, while the trend of useful rate and waste rate are inversely related. Concerned policymakers can set different thresholds dependent up the number and budgetary constraints of police forces, and they can work towards achieving crime prevention in urban hotspots. Importantly, Poisson’s approach can be simply implemented with Excel, be conducive to drive by the office practitioner, and elevate the transparency of crime prediction.