Explore the vast range of titles available.

We sought to summarize the study design, modelling strategies, and performance measures reported in studies on clinical prediction models developed using machine learning techniques. We search PubMed for articles published between 01/01/2018 and 31/12/2019, describing the development or the development with external validation of a multivariable prediction model using any supervised machine learning technique. No restrictions were made based on study design, data source, or predicted patient-related health outcomes. We included 152 studies, 58 (38.2% [95% CI 30.8–46.1]) were diagnostic and 94 (61.8% [95% CI 53.9–69.2]) prognostic studies. Most studies reported only the development of prediction models (n = 133, 87.5% [95% CI 81.3–91.8]), focused on binary outcomes (n = 131, 86.2% [95% CI 79.8–90.8), and did not report a sample size calculation (n = 125, 82.2% [95% CI 75.4–87.5]). The most common algorithms used were support vector machine (n = 86/522, 16.5% [95% CI 13.5–19.9]) and random forest (n = 73/522, 14% [95% CI 11.3–17.2]). Values for area under the Receiver Operating Characteristic curve ranged from 0.45 to 1.00. Calibration metrics were often missed (n = 494/522, 94.6% [95% CI 92.4–96.3]). Our review revealed that focus is required on handling of missing values, methods for internal validation, and reporting of calibration to improve the methodological conduct of studies on machine learning–based prediction models. PROSPERO, CRD42019161764.

A recurrent concern about machine learning algorithms is that they operate as “black boxes,” making it difficult to identify how and why the algorithms reach particular decisions, recommendations, or predictions. Yet judges are confronting machine learning algorithms with increasing frequency, including in criminal, administrative, and civil cases. This Essay argues that judges should demand explanations for these algorithmic outcomes. One way to address the “black box” problem is to design systems that explain how the algorithms reach their conclusions or predictions. If and as judges demand these explanations, they will play a seminal role in shaping the nature and form of “explainable AI” (xAI). Using the tools of the common law, courts can develop what xAI should mean in different legal contexts. There are advantages to having courts to play this role: Judicial reasoning that builds from the bottom up, using case-by-case consideration of the facts to produce nuanced decisions, is a pragmatic way to develop rules for xAI. Further, courts are likely to stimulate the production of different forms of xAI that are responsive to distinct legal settings and audiences. More generally, we should favor the greater involvement of public actors in shaping xAI, which to date has largely been left in private hands.

BackgroundStress echocardiography is a well-established diagnostic prognostic modality with good predictive power in patient with suspected coronary artery disease (CAD). Cardiovascular risk factors are used in the assessment of the probability of CAD. The link between the outcome of stress echocardiography and patients’ variables has not been widely investigated.ObjectiveThe aim of this study aim is to understand whether new algorithm can be developed using machine learning principle to predict stress echocardiography result in patients with suspected CAD based on clinical variables. This could facilitate a reclassification of patients’ risk prediction of CAD.MethodsA machine learning framework was generated to automate the prediction of inducible ischemia on stress echocardiography results. The framework consisted of four stages: feature extraction, pre-processing, feature selection, and classification stage. A mutual information‘based feature selection method was used to investigate the amount of information that each feature carried to define the positive outcome of stress echocardiography. Two classification algorithms, support vector machine (SVM) and random forest classifiers, have been deployed.Data from 2201 patients were used to train and validate the framework. Patients mean age was 62 (SD 11) years. The data consists of anthropological data and cardiovascular risk factors such as gender, age, weight, family history, diabetes, smoking history, hypertension, hypercholesterolemia, atrial fibrillation, prior diagnosis of CAD, prior bypass grafting, history of coronary intervention, stroke, airways disease, chronic inflammatory disease, chronic kidney disease (>3) and prescribed medications at the time of the test. There were 394 positive and 1807 negative stress echocardiography results. The framework was evaluated using the whole dataset including cases with prior diagnosis of CAD. Five-fold cross-validation was used to validate the performance of the framework. We also investigated the model in the subset of patients with no prior CAD.ResultsThe feature selection methods showed that prescribed medications such as antiplatelet and angiotensin-converting enzyme inhibitor, weight, and diabetes were the features that shared the most information about the outcome of stress echocardiography. Random forest classifiers showed the best trade-off between sensitivity and specificity and was achieved with three features. Using only four features, we achieved an accuracy of 82.13%.ConclusionsThis study shows that machine learning can predict the outcome of stress echocardiography based on only a few features. Further research correlating the clinical variable and stress echo result to cardiovascular outcomes improve the performance of the proposed algorithm with the potential of facilitating patient selection for early treatment/intervention avoiding unnecessary downstream testing.Conflict of Interestnone

Objects that possess mass (e.g., automobiles, manufactured items, etc.) translationally accelerate in direct proportion to the force applied scaled by the object’s mass in accordance with Newton’s Law, while the rotational companion is Euler’s moment equations relating angular acceleration of objects that possess mass moments of inertia. Michel Chasles’s theorem allows us to simply invoke Newton and Euler’s equations to fully describe the six degrees of freedom of mechanical motion. Many options are available to control the motion of objects by controlling the applied force and moment. A long, distinguished list of references has matured the field of controlling a mechanical motion, which culminates in the burgeoning field of deterministic artificial intelligence as a natural progression of the laudable goal of adaptive and/or model predictive controllers that can be proven to be optimal subsequent to their development. Deterministic A.I. uses Chasle’s claim to assert Newton’s and Euler’s relations as deterministic self-awareness statements that are optimal with respect to state errors. Predictive controllers (both continuous and sampled-data) derived from the outset to be optimal by first solving an optimization problem with the governing dynamic equations of motion lead to several controllers (including a controller that twice invokes optimization to formulate robust, predictive control). These controllers are compared to each other with noise and modeling errors, and the many figures of merit are used: tracking error and rate error deviations and means, in addition to total mean cost. Robustness is evaluated using Monte Carlo analysis where plant parameters are randomly assumed to be incorrectly modeled. Six instances of controllers are compared against these methods and interpretations, which allow engineers to select a tailored control for their given circumstances. Novel versions of the ubiquitous classical proportional-derivative, “PD” controller, is developed from the optimization statement at the outset by using a novel re-parameterization of the optimal results from time-to-state parameterization. Furthermore, time-optimal controllers, continuous predictive controllers, and sampled-data predictive controllers, as well as combined feedforward plus feedback controllers, and the two degree of freedom controllers (i.e., 2DOF). The context of the term “feedforward” used in this study is the context of deterministic artificial intelligence, where analytic self-awareness statements are strictly determined by the governing physics (of mechanics in this case, e.g., Chasle, Newton, and Euler). When feedforward is combined with feedback per the previously mentioned method (provenance foremost in optimization), the combination is referred to as “2DOF” or two degrees of freedom to indicate the twice invocation of optimization at the genesis of the feedforward and the feedback, respectively. The feedforward plus feedback case is augmented by an online (real time) comparison to the optimal case. This manuscript compares these many optional control strategies against each other. Nominal plants are used, but the addition of plant noise reveals the robustness of each controller, even without optimally rejecting assumed-Gaussian noise (e.g., via the Kalman filter). In other words, noise terms are intentionally left unaddressed in the problem formulation to evaluate the robustness of the proposed method when the real-world noise is added. Lastly, mismodeled plants controlled by each strategy reveal relative performance. Well-anticipated results include the lowest cost, which is achieved by the optimal controller (with very poor robustness), while low mean errors and deviations are achieved by the classical controllers (at the highest cost). Both continuous predictive control and sampled-data predictive control perform well at both cost as well as errors and deviations, while the 2DOF controller performance was the best overall.

Child undernutrition is a global public health problem with serious implications. In this study, we estimate predictive algorithms for the determinants of childhood stunting by using various machine learning (ML) algorithms. This study draws on data from the Ethiopian Demographic and Health Survey of 2016. Five ML algorithms including eXtreme gradient boosting, k-nearest neighbours (k-NN), random forest, neural network and the generalised linear models were considered to predict the socio-demographic risk factors for undernutrition in Ethiopia. Households in Ethiopia. A total of 9471 children below 5 years of age participated in this study. The descriptive results show substantial regional variations in child stunting, wasting and underweight in Ethiopia. Also, among the five ML algorithms, xgbTree algorithm shows a better prediction ability than the generalised linear mixed algorithm. The best predicting algorithm (xgbTree) shows diverse important predictors of undernutrition across the three outcomes which include time to water source, anaemia history, child age greater than 30 months, small birth size and maternal underweight, among others. The xgbTree algorithm was a reasonably superior ML algorithm for predicting childhood undernutrition in Ethiopia compared to other ML algorithms considered in this study. The findings support improvement in access to water supply, food security and fertility regulation, among others, in the quest to considerably improve childhood nutrition in Ethiopia.

As applications like electric vehicles, all-electric ships, and all-electric aircraft continue to evolve, Noise, Vibration, and Harshness (NVH) issues have garnered extensive attention. However, as the core of the power system, permanent magnet synchronous motors (PMSMs) still lack control algorithms that consider vibration problems. Therefore, this paper proposes a model-based predictive vibration suppression algorithm to suppress the PMSM vibration. Firstly, this paper explores the influence of armature currents on vibration by analyzing the vibration characteristics of PMSMs, and proposes a minimum vibration current model. On this basis, according to the torque conditions required for the stable operation of the motor, a model-based predictive vibration suppression algorithm is designed. Finally, the effectiveness of the proposed algorithm is verified through prototype experiments.

Dissolved oxygen (DO) concentration in water is one of the key parameters for assessing river water quality. Artificial intelligence (AI) methods have previously proved to be accurate tools for DO concentration prediction. This study presents the implementation of a deep learning approach applied to a recurrent neural network (RNN) algorithm. The proposed deep recurrent neural network (DRNN) model is compared with support vector machine (SVM) and artificial neural network (ANN) models, formerly shown to be robust AI algorithms. The Fanno Creek in Oregon (USA) is selected as a case study and daily values of water temperature, specific conductance, streamflow discharge, pH, and DO concentration are used as input variables to predict DO concentration for three different lead times (“t + 1,” “t + 3,” and “t + 7”). Based on Pearson’s correlation coefficient, several input variable combinations are formed and used for prediction. The model prediction performance is evaluated using various indices such as correlation coefficient, Nash–Sutcliffe efficiency, root mean square error, and mean absolute error. The results identify the DRNN model ( CC Testing = 0.97 , N S E Testing = 0.948 , RMSE Testing = 0.43 and MAE Testing = 0.25 ) as the most accurate among the three models considered, highlighting the potential of deep learning approaches for water quality parameter prediction.

Background Coronary heart disease (CHD), the most common form of heart disease, progresses over years before culminating in serious cardiac events. Early prediction and intervention are critical to reducing CHD-related morbidity, mortality, and healthcare burden. Objective To develop and validate a machine learning model using statewide electronic health records (EHRs) to predict 1-year risk of CHD in the general population of Maine, enabling targeted preventive strategies. Methods Two population-based cohorts were constructed from the Maine Health Information Exchange (HIE): a retrospective cohort for model training and calibration (2015–2017, N  = 1,042,124), and a prospective cohort for external validation (2016–2018, N  = 1,040,158). EHR features included demographics, diagnoses, procedures, medications, labs, and utilization metrics. A multistage modeling pipeline—comprising statistical filtering, XGBoost-based feature selection, risk prediction, and isotonic regression calibration—was used to construct the final model. Validation included discrimination, calibration, and survival analysis. Results The final XGBoost model achieved strong discrimination: AUC = 0.952 (95% CI: 0.950–0.954) in the retrospective cohort and 0.888 (95% CI: 0.885–0.890) in the prospective cohort. Based on calibrated risk probabilities, the population was stratified into five risk categories: very low (92.30%, N  = 960,021), low (6.79%, N  = 70,676), medium (0.85%, N  = 8,888), high (0.05%, N  = 554), and very high (0.002%, N  = 19). Among the very high-risk group, 11 individuals (57.89%) developed CHD within one year. Conclusions This statewide, HIE-based CHD risk prediction model demonstrates robust performance and real-world applicability. It enables early identification of high-risk individuals and supports population-scale precision prevention through evidence-informed, proactive care.

Abstract Context Primary aldosteronism (PA) affects 10% to 15% of individuals with hypertension and increases cardiovascular risk. Differentiating between unilateral and bilateral PA determines optimal treatment and typically requires adrenal vein sampling (AVS). Emerging subtyping methods include predictive algorithms and nuclear imaging. Objective This study explores hypertensive individuals' preferences for different PA subtyping strategies. Methods Two labeled discrete choice experiments (DCEs) evaluated preferences for subtyping methods based on test accuracy, waiting time, adverse effects, and out-of-pocket cost. Latent class conditional logit (LCL) modeling segmented participants by preferences, while policy simulation analyses examined uptake variations by age, sex, and income. Results Among 583 hypertensive Australian adults (mean age: 48 years; 48% female), 85% were willing to undergo PA subtyping. Participants prioritized accuracy, shorter waiting times, minimal side effects, and lower costs. LCL analysis revealed that participants who were older, female, or considered themselves less busy were more likely to opt for PA subtyping. Subtyping uptake was highest for algorithm-based methods (∼54%) with its uptake rate increasing to 68% after factoring in cost. Conclusion Preferences for PA subtyping are driven by cost, invasiveness, and waiting time. Less-invasive, faster, and low-cost methods were preferred, even if they are slightly less accurate than AVS. Further research is needed to optimize the accuracy of subtyping algorithms and facilitate implementation in clinical practice.

Background When developing or updating a predictive algorithm to make predictions (e.g., risk estimates) in individuals, the sample size of the development (training) dataset is an important consideration. Small datasets reduce a model’s predictive performance, generalisability and fairness, and may lead to harm. To help identify (minimum) sample sizes required for model development or updating, various sample size calculations exist but their underlying theory is based on standard (unpenalised) regression and focus on minimising overfitting. To address this, we propose a more general approach, which allows extension to any machine learning or AI-based modelling method, and to any performance metric (estimand) of interest. Methods The proposed approach draws samples from anticipated posterior distributions to examine the impact on degradation in a model’s predictive performance compared to a reference model. It requires researchers to provide candidate predictors (features), a reference model (e.g., based on mean outcome incidence, predictor weights and c-statistics of previous models), and a (existing, synthetic or pilot) dataset reflecting the joint distribution of candidate predictors in the target population. A fully simulation-based approach then generates thousands of models conditional on a chosen sample size and model development strategy, to produce posterior distributions of individual predictions and model performance (degradation) metrics, to inform required sample sizes. To improve computational speed for penalised regression, we also propose a one-sample Bayesian approximation that combines shrinkage priors with a likelihood decomposed into sample size and Fisher’s unit information. Results The approaches are illustrated when developing models for pre-eclampsia using logistic regression (unpenalised, uniform shrinkage, lasso or ridge) and random forests; and are shown to encompass existing sample size calculation criteria whilst also providing model assurance probabilities, instability metrics, and degradation statistics about calibration, discrimination, clinical utility, prediction error and fairness. Uncertainty in the reference model can also be accounted for in the calculations. Conclusions Our approach provides a general framework for (minimum) sample size calculations for developing or updating a prediction model, and is applicable for any statistical, machine learning or AI-based development method for supervised learning. Crucially, the recommended (minimum) sample size will depend on users’ chosen estimands and model development (or updating) approach. Example R and Stata code is provided.