Explore the vast range of titles available.

Abstract Study Objectives If properly consumed, caffeine can safely and effectively mitigate the effects of sleep loss on alertness. However, there are no tools to determine the amount and time to consume caffeine to maximize its effectiveness. Here, we extended the capabilities of the 2B-Alert app, a unique smartphone application that learns an individual’s trait-like response to sleep loss, to provide personalized caffeine recommendations to optimize alertness. Methods We prospectively validated 2B-Alert’s capabilities in a 62-hour total sleep deprivation study in which 21 participants used the app to measure their alertness throughout the study via the psychomotor vigilance test (PVT). Using PVT data collected during the first 36 hours of the sleep challenge, the app learned the participant’s sleep-loss response and provided personalized caffeine recommendations so that each participant would sustain alertness at a pre-specified target level (mean response time of 270 milliseconds) during a 6-hour period starting at 44 hours of wakefulness, using the least amount of caffeine possible. Starting at 42 hours, participants consumed 0 to 800 mg of caffeine, per the app recommendation. Results 2B-Alert recommended no caffeine to five participants, 100–400 mg to 11 participants, and 500–800 mg to five participants. Regardless of the consumed amount, participants sustained the target alertness level ~80% of the time. Conclusions 2B-Alert automatically learns an individual’s phenotype and provides personalized caffeine recommendations in real time so that individuals achieve a desired alertness level regardless of their sleep-loss susceptibility. Graphical abstract Graphical Abstract

In retail operations, customer choices may be affected by stockout and promotion events. Given panel data with the transaction history of customers, and product availability and promotion data, our goal is to predict future individual purchases. We use a general nonparametric framework in which we represent customers by partial orders of preferences. In each store visit, each customer samples a full preference list of the products consistent with her partial order, forms a consideration set, and then chooses to purchase the most preferred product among the considered ones. Our approach involves: (a) defining behavioral models to build consideration sets as subsets of the products on offer, (b) proposing a clustering algorithm for determining customer segments, and (c) deriving marginal distributions for partial preferences under the multinomial logit model. Numerical experiments on real-world panel data show that our approach allows more accurate, fine-grained predictions for individual purchase behavior compared to state-of-the-art alternative methods. The online appendix is available at https://doi.org/10.1287/mnsc.2016.2683 . This paper was accepted by Vishal Gaur, operations management.

BACKGROUND:Cystic fibrosis (CF) is an inherited, chronic, progressive condition affecting around 10,000 individuals in the United Kingdom and over 70,000 worldwide. Survival in CF has improved considerably over recent decades, and it is important to provide up-to-date information on patient prognosis. METHODS:The UK Cystic Fibrosis Registry is a secure centralized database, which collects annual data on almost all CF patients in the United Kingdom. Data from 43,592 annual records from 2005 to 2015 on 6181 individuals were used to develop a dynamic survival prediction model that provides personalized estimates of survival probabilities given a patient’s current health status using 16 predictors. We developed the model using the landmarking approach, giving predicted survival curves up to 10 years from 18 to 50 years of age. We compared several models using cross-validation. RESULTS:The final model has good discrimination (C-indexes0.873, 0.843, and 0.804 for 2-, 5-, and 10-year survival prediction) and low prediction error (Brier scores0.036, 0.076, and 0.133). It identifies individuals at low and high risk of short- and long-term mortality based on their current status. For patients 20 years of age during 2013–2015, for example, over 80% had a greater than 95% probability of 2-year survival and 40% were predicted to survive 10 years or more. CONCLUSIONS:Dynamic personalized prediction models can guide treatment decisions and provide personalized information for patients. Our application illustrates the utility of the landmarking approach for making the best use of longitudinal and survival data and shows how models can be defined and compared in terms of predictive performance.

In recent years, there has been a growing demand to develop efficient recommender systems which track users' preferences and recommend potential items of interest to users. In this article, we propose a group-specific method to use dependency information from users and items which share similar characteristics under the singular value decomposition framework. The new approach is effective for the \"cold-start\" problem, where, in the testing set, majority responses are obtained from new users or for new items, and their preference information is not available from the training set. One advantage of the proposed model is that we are able to incorporate information from the missing mechanism and group-specific features through clustering based on the numbers of ratings from each user and other variables associated with missing patterns. In addition, since this type of data involves large-scale customer records, traditional algorithms are not computationally scalable. To implement the proposed method, we propose a new algorithm that embeds a back-fitting algorithm into alternating least squares, which avoids large matrices operation and big memory storage, and therefore makes it feasible to achieve scalable computing. Our simulation studies and MovieLens data analysis both indicate that the proposed group-specific method improves prediction accuracy significantly compared to existing competitive recommender system approaches. Supplementary materials for this article are available online.

Freezing of gait (FoG) is a common severe gait disorder in patients with advanced Parkinson’s disease. The ability to predict the onset of FoG episodes early on allows for timely intervention, which is essential for improving the life quality of patients. Machine learning and deep learning, the current methods, face real-time diagnosis challenges due to comprehensive data processing requirements. Their “black box” nature makes interpreting features and classification boundaries difficult. In this manuscript, we explored a dynamic mode decomposition (DMD)-based approach together with optimal delay embedding time to reconstruct and predict the time evolution of acceleration signals, and introduced a triple index based on DMD to predict and classify FoG. Our predictive analysis shows 86.5% accuracy in classification, and an early prediction ratio of 81.97% with an average early prediction time of 6.13 s. This DMD-based approach has the potential for real-time patient-specific FoG prediction.

We integrated molecular data with available prognostic factors in patients undergoing allogeneic hematopoietic cell transplantation (alloHCT) for myelodysplastic syndrome (MDS) or secondary acute myeloid leukemia (sAML) from MDS to evaluate their impact on prognosis. Three hundred four patients were sequenced for mutations in 54 genes. We used a Cox multivariate model and competing risk analysis with internal and cross validation to identify factors prognostic of overall survival (OS), cumulative incidence of relapse (CIR), and non-relapse mortality (NRM). In multivariate analysis, mutated NRAS , U2AF1 , IDH2 , and TP53 and/or a complex karyotype were significant prognostic markers for OS besides age above 60 years, remission status, IPSS-R cytogenetic risk, HCT-CI > 2 and female donor sex. Mutated NRAS , IDH1 , EZH2 , and TP53 and/or a complex karyotype were genetic aberrations with prognostic impact on CIR. No molecular markers were associated with the risk of NRM. The inclusion of molecular information results in better risk prediction models for OS and CIR when assessed by the Akaike information criterion. Internal cross validation confirmed the robustness of our comprehensive risk model. In summary, we propose to combine molecular, cytogenetic, and patient- and transplantation-associated risk factors into a comprehensive risk model to provide personalized predictions of outcome after alloHCT.

Personalised prediction of functional outcomes is a promising approach for targeted early intervention in psychiatry. However, generalisability and resource efficiency of such prognostic models represent challenges. In the PRONIA study (German Clinical Trials Register: DRKS00005042), we demonstrate excellent generalisability of prognostic models in individuals at clinical high-risk for psychosis or with recent-onset depression, and substantial contributions of detailed clinical phenotyping, particularly to the prediction of role functioning. These results indicate that it is possible that functioning prediction models based only on clinical data could be effectively applied in diverse healthcare settings, so that neuroimaging data may not be needed at early assessment stages.

Background Prognostic information for patients with hypertension is largely based on population averages. The purpose of this study was to compare the performance of four machine learning approaches for personalized prediction of incident hospitalization for cardiovascular disease among newly diagnosed hypertensive patients. Methods Using province-wide linked administrative health data in Alberta, we analyzed a cohort of 259,873 newly-diagnosed hypertensive patients from 2009 to 2015 who collectively had 11,863 incident hospitalizations for heart failure, myocardial infarction, and stroke. Linear multi-task logistic regression, neural multi-task logistic regression, random survival forest and Cox proportional hazard models were used to determine the number of event-free survivors at each time-point and to construct individual event-free survival probability curves. The predictive performance was evaluated by root mean squared error, mean absolute error, concordance index, and the Brier score. Results The random survival forest model has the lowest root mean squared error value at 33.94 and lowest mean absolute error value at 28.37. Machine learning methods provide similar discrimination and calibration in the personalized survival prediction of hospitalizations for cardiovascular events in patients with hypertension. Neural multi-task logistic regression model has the highest concordance index at 0.8149 and lowest Brier score at 0.0242 for the personalized survival prediction. Conclusions This is the first personalized survival prediction for cardiovascular diseases among hypertensive patients using administrative data. The four models tested in this analysis exhibited a similar discrimination and calibration ability in predicting personalized survival prediction of hypertension patients.

Azacytidine is the only approved treatment for patients with higher-risk myelodysplastic syndromes (MDS); yet less than half of the patients will achieve a response, whereas the duration of response is highly heterogeneous and there are no predictors for response duration. The aim of this study is to estimate the patient’s time to loss of response (LoR) to azacytidine based on clinical measurements during treatment. To this end, a personalized prediction framework is proposed that estimates the LoR of a new patient using a patient similarity-based approach. Namely, the new patient’s clinical data—represented as a multivariate time series—are compared to a reference set of patients. The comparison uses distance metrics that quantify how similar two patients’ time series are, assuming patients with similar trajectories tend to have similar LoR. Then, the LoR of the new patient is predicted by averaging the outcomes of the most similar reference patients. The pipeline includes a data normalization strategy that centers each feature on its baseline value and scales it to highlight relative changes and distance metrics to quantify similarity. Both real-world and simulated data were utilized to evaluate the proposed methodology, employing the leave-one-out validation and the Mean Absolute Percentage Error (MAPE) to assess accuracy. The estimated MAPE was found to be 30.52% and 11.82% in the real-world and simulated dataset, respectively. The best and most robust predictions were achieved using the Euclidean distance metric and setting the number of most similar patients around three to five. This study proposes a personalized predictive approach for the LoR to azacitidine in the MDS clinical setting, demonstrating potential for a serviceable prediction of LoR and forming the foundation for further research.

Background There is a great need for the development of personalized prediction models (PPMs) that can predict the rate of disease progression for persons with Parkinson's disease (PD), based on their individual characteristics. In this study, we aimed to clarify the perspective of persons diagnosed with PD on the value of such hypothetical PPMs. Methods We organized four focus group discussions, each including five persons with PD who were diagnosed within the last 5 years. The sessions focused on what they think of receiving a personalized prediction; what outcomes are important to them; if and how the possibility of influencing the prognosis would change the way they think of personalized predictions; how they deal with the uncertainty from a PPM; and what barriers and facilitators they expect for using a PPM. Results The wish of persons with PD for receiving personalized prognostic information was highly heterogenous, for various reasons. Most persons with PD would like to receive more personalized prognostic information, mainly to better prepare themselves and their loved ones for the future. The prediction provided should be as personalized as possible, and there should be adequate supervision and coaching by a professional when providing the information. They were particularly interested in receiving prognostic information when their interventions would be available that could subsequently influence the identified prognostic factor and thereby affect the disease course beneficially. Conclusion Most persons with PD in this study want more insight into their own future by means of prediction models, provided that this is explained and supervized properly by professionals. Patient or Public Contribution Two patient‐researchers were involved in the study design, conduct of the study, interpretation of the data and in preparation of the manuscript.