scholarly journals Identifying patients at high risk of inappropriate drug dosing in periods with renal dysfunction

2021 ◽  
Author(s):  
Benjamin Skov Kaas-Hansen ◽  
Cristina Leal Rodríguez ◽  
Davide Placido ◽  
Hans-Christian Thorsen-Meyer ◽  
Anna Pors Nielsen ◽  
...  
Keyword(s):  
Neural Network ◽  
High Risk ◽  
Kidney Failure ◽  
Model Development ◽  
Real Life ◽  
Inappropriate Drug ◽  
Prospective Clinical Study ◽  
Hospital Data ◽  
Neural Network Models ◽  
Index Set

Introduction: Dosing of renally cleared drugs in patients with kidney failure often deviates from clinical guidelines but little is known about what is predictive of receiving inappropriate doses. Methods and materials: We combined data from the Danish National Patient Register and in-hospital data on drug administrations and estimated glomerular filtration rates for admissions between 1 October 2009 and 1 June 2016, from a pool of about 2.9 million persons. We trained artificial neural network and linear logistic ridge regression models to predict the risk of five outcomes (>0, ≥1, ≥2, ≥3 and ≥5 inappropriate doses daily) with index set 24 hours after admission. We used time-series validation for evaluating discrimination, calibration, clinical utility and explanations. Results: Of 52,451 admissions included, 42,250 (81%) were used for model development. The median age was 77 years; 50% of admissions were of women. ≥5 drugs were used between admission start and index in 23,124 admissions (44%); the most common drug classes were analgesics, systemic antibacterials, diuretics, antithrombotics, and antacids. The neural network models had better discriminative power (all AUROCs between 0.77 and 0.81) and were better calibrated than their linear counterparts. The main prediction drivers were use of anti-inflammatory, antidiabetic and anti-Parkison's drugs as well as having a diagnosis of chronic kidney failure. Sex and age affected predictions but slightly. Conclusion: Our models can flag patients at high risk of receiving at least one inappropriate dose daily in a controlled in-silico setting. A prospective clinical study may confirm this holds in real-life settings and translates into benefits in hard endpoints.

scholarly journals Legitimising neural network river forecasting models: a new data-driven mechanistic modelling framework

2013 ◽  
Vol 10 (1) ◽  
pp. 145-187 ◽  
Author(s):  
N. J. Mount ◽  
C. W. Dawson ◽  
R. J. Abrahart
Keyword(s):  
Neural Network ◽  
Goodness Of Fit ◽  
Model Development ◽  
Relative Sensitivity ◽  
Data Driven ◽  
Neural Network Models ◽  
Core Element ◽  
Mechanistic Modelling ◽  
Mechanistic Understanding ◽  
River Forecasting

Abstract. In this paper we address the difficult problem of gaining an internal, mechanistic understanding of a neural network river forecasting (NNRF) model. Neural network models in hydrology have long been criticised for their black-box character, which prohibits adequate understanding of their modelling mechanisms and has limited their broad acceptance by hydrologists. In response, we here present a new, data-driven mechanistic modelling (DDMM) framework that incorporates an evaluation of the legitimacy of a neural network's internal modelling mechanism as a core element in the model development process. The framework is exemplified for two NNRF modelling scenarios, and uses a novel adaptation of first order, partial derivate, relative sensitivity analysis methods as the means by which each model's mechanistic legitimacy is explored. The results demonstrate the limitations of standard, goodness-of-fit validation procedures applied by NNRF modellers, by highlighting how the internal mechanisms of complex models that produce the best fit scores can have much lower legitimacy than simpler counterparts whose scores are only slightly inferior. The study emphasises the urgent need for better mechanistic understanding of neural network-based hydrological models and the further development of methods for elucidating their mechanisms.

scholarly journals Enhanced prediction of atrial fibrillation and mortality among patients with congenital heart disease using nationwide register based medical hospital data and neural networks

2021 ◽  
Author(s):  
Kok Wai Giang ◽  
Saga Helgadottir ◽  
Mikael Dellborg ◽  
Giovanni Volpe ◽  
Zacharias Mandalenakis
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Neural Networks ◽  
Congenital Heart Disease ◽  
Logistic Regression ◽  
Heart Disease ◽  
Congenital Heart ◽  
Hospital Data ◽  
Neural Network Models ◽  
Over Time

Abstract Aims To improve short-and long-term predictions of mortality and atrial fibrillation among patients with congenital heart disease from a nationwide population using neural networks. Methods and results The Swedish National Patient Register and the Cause of Death Register were used to identify all patients with congenital heart disease born from 1970 to 2017. A total of 71,941 congenital heart disease patients were identified and followed-up from birth until the event or end of study in 2017. Based on data from a nationwide population, a neural network model was obtained to predict mortality and atrial fibrillation. Logistic regression based on the same data was used as a baseline comparison. Of 71,941 congenital heart disease patients, a total of 5768 died (8.02%) and 995 (1.38%) developed atrial fibrillation over time with a mean follow-up time of 16.47 years (standard deviation 12.73 years). The performance of neural network models in predicting the mortality and atrial fibrillation was higher than the performance of logistic regression regardless of the complexity of the disease, with an average Area Under the Receiver Operating Characteristic of > 0.80 and >0.70, respectively. The largest differences were observed in mortality and complexity of congenital heart disease over time. Conclusion We found that neural networks can be used to predict mortality and atrial fibrillation on a nationwide scale using data that are easily obtainable by clinicians. In addition, neural networks showed a high performance overall and, in most cases, with better performance for prediction as compared with more traditional regression methods.

A STUDY OF EARLY STOPPING AND MODEL SELECTION APPLIED TO THE PAPERMAKING INDUSTRY

2000 ◽  
Vol 10 (01) ◽  
pp. 9-18 ◽  
Author(s):  
PETER J. EDWARDS ◽  
ALAN F. MURRAY
Keyword(s):  
Neural Network ◽  
Model Selection ◽  
Cross Validation ◽  
Model Development ◽  
Network Models ◽  
Comparison Study ◽  
Early Stopping ◽  
Neural Network Models ◽  
Model Evidence ◽  
Fold Cross Validation

This paper addresses the issues of neural network model development and maintenance in the context of a complex task taken from the papermaking industry. In particular, it describes a comparison study of early stopping techniques and model selection, both to optimise neural network models for generalisation performance. The results presented here show that early stopping via use of a Bayesian model evidence measure is a viable way of optimising performance while also making maximum use of all the data. In addition, they show that ten-fold cross-validation performs well as a model selector and as an estimator of prediction accuracy. These results are important in that they show how neural network models may be optimally trained and selected for highly complex industrial tasks where the data are noisy and limited in number.

scholarly journals P0184RISK PREDICTION FOR DEATH AND END-STAGE RENAL DISEASE DOES NOT PARALLEL REAL-LIFE TRAJECTORY OF OLDER PATIENTS WITH ADVANCED CHRONIC KIDNEY DISEASE - A ROMANIAN CENTER EXPERIENCE

2020 ◽  
Vol 35 (Supplement_3) ◽  
Author(s):  
Andra Nastasa ◽  
Mugurel Apetrii ◽  
Mihai Onofriescu ◽  
Ionut Nistor ◽  
Hani Hussien ◽  
...  
Keyword(s):  
High Risk ◽  
Mortality Risk ◽  
Kidney Failure ◽  
Older Patients ◽  
Real Life ◽  
Low Risk ◽  
Individual Risk ◽  
Failure Risk ◽  
Risk Of Mortality ◽  
Life Trajectory

Abstract Background and Aims In Europe, the share of the elderly (≥65 years of age) in the total population is estimated to increase from 19.2% in 2016 to 29.1% by 2080. In 2016, European Renal Best Practice (ERBP) group published a clinical practice guideline on management of older patients with CKD stage3b or higher (eGFR<45ml/min/1.73 m2). Two risk stratifications scores were emphasized: Bansal score for prognosticating risk of death in medium term, and Kidney Failure Risk Equation (KFRE) for estimating progression of CKD stage 3b or 4 to ESRD. Our group, as part of the ERBP team, aimed to evaluate and apply the framework proposed by the guideline, consisting of risk prediction for both mortality and progression to ESRD in a cohort of elderly patients with advanced CKD. After dividing the population in groups of risk, we described their real-life trajectory in terms of either reaching ESRD/death. Method In this retrospective cohort study we included patients aged ≥65 years with CKD stage 3b-4, evaluated at the Outpatient Nephrology Department of Dr. C. I. Parhon Hospital from Iași, Romania, between October 2016 – October 2018. Individual risk for mortality was predicted using Bansal score, a nine-variable equation model. A total score of 7 (associated with a mortality risk of 53.82%) was established as cut-off value to differentiate between 2 groups: high risk of mortality (Bansal ≥ 7) and low risk of mortality (Bansal < 7), given the fact that the ERBP guidelines don’t define a threshold for high risk in respect to mortality outcome. According to the algorithm proposed by the guideline, individual risk for progression to ESRD at 5 years was calculated in the low mortality risk group, using the 4-variable Kidney Failure Risk Equation (KFRE). Results The final cohort included 958 patients, with a mean age of 74 years (SD: 7), and with similar gender distribution (50.6% female vs. 49.4% male). Predicted trajectory in terms of reaching ESRD / death: When we applied Bansal score for mortality, the total study population (N=958) was divided in two groups: N1 with high risk of mortality, which comprised more than half of the cohort (548 patients, 57.2%) and N2 with low risk of mortality (410 patients, 42.8%). Individual risk of progression to ESRD was then estimated in N2 group, using 4-variable KFRE. Nearly ¾ of this group (75.4%, 309 subjects) presented a low-risk of progression and ¼ (24.6%, 101 subjects) had high-risk. Real-life trajectory in terms of reaching ESRD / death: From the entire cohort, 31 patients started renal replacement therapy (RRT) and 164 patients died as their first clinical event. The RRT initiation rate was 3.6% of N1 group (20 subjects) versus 2.7% of N2 group (11 subjects). The mortality rate was 15.5% of N1 group (85 deaths) versus 19.3% of N2 group (79 deaths). Figure 1 depicts the real-life trajectory of the population groups in terms of reaching ESRD / death. Conclusion In a large population from Eastern Europe, the application of the algorithm from the Clinical Practice Guideline on management of older patients with advanced CKD showed that risk prediction for death and end-stage renal disease does not parallel the real-life trajectory of the population. More than half of the subjects had a high risk of mortality, however we found similar death rates in the 2 groups (high versus low risk of mortality). Also, the RRT initiation rates were similar, irrespective of predicted mortality risk or kidney failure risk, suggesting that implementing the guideline in real-life settings is still a challenge.

scholarly journals ВЫБОР МОДЕЛИ БИОЛОГИЧЕСКОЙ НЕЙРОННОЙ СЕТИ ДЛЯ СЕГМЕНТАЦИИ ИЗОБРАЖЕНИЯ БИОЖИДКОСТНОЙ ПОВЕРХНОСТИ

2019 ◽  
pp. 78-93
Author(s):  
М.Е. Семенов ◽  
Т.Ю. Заблоцкая
Keyword(s):  
Neural Network ◽  
Pattern Recognition ◽  
Network Architecture ◽  
Real Life ◽  
Image Data ◽  
Network Models ◽  
Digital Processing ◽  
Neural Network Models ◽  
Biological Neural Network ◽  
Delay Differential

In the paper, the biological neural network models are analyzed with a purpose to solve the problems of segmentation and pattern recognition when applied to the bio-liquid facies obtained by the cuneiform dehydration method. The peculiarities of the facies’ patterns and the key steps of their digital processing are specified in the frame of the pattern recognition. Feasibility of neural network techniques for the different image data level digital processing is reviewed as well as for image segmentation. The real-life biological neural network architecture concept is described using the mechanisms of the electrical input-output membrane voltage and both induced and endogenic (spontaneous) activities of the neural clusters when spiking. The mechanism of spike initiation is described for metabotropic and ionotropic receptive clusters with the nature of environmental exciting impact specified. Also, the mathematical models of biological neural networks that comprise ot only functional nonlinearities but the hysteretic ones are analyzed and the reasons are given for preference of the mathematical model with delay differential equations is chosen providing its applicability for modeling a single neuron and neural network as well. В работе рассматривается применение моделей биологической нейронной сети для сегментации изображения фации биожидкости, полученной методом клиновидной дегидратации. Выделены основные характерные особенности, присущие паттернам фаций биожидкостей, а также основные этапы их цифровой обработки в рамках задачи распознавания образов. Проведен анализ использования искусственных нейронных сетей для цифровой обработки изображений для разных уровней представления данных; сделан обзор основных нейросетевых методов сегментации. Описан принцип построения биологически достоверных искусственных нейронных сетей, использующих механизмы изменения мембранного потенциала нейронов и учитывающих при генерации спайка как вызванную активность, так и эндогенную (спонтанную) активность нейронных кластеров. Описан механизм инициации спайка для метаботропных и ионотропных рецептивных кластеров с указанием природы запускающего внешнего воздействия. Проведен анализ существующих математических моделей биологических нейросетей, содержащих помимо обычных функциональных нелинейностей нелинейности гистерезисной природы. Сделан выбор в пользу математической модели, использующей дифференциальные уравнения с запаздыванием, которые могут быть применены как для описания отдельного биологического нейрона, так и для описания работы нейронной сети.

Trigonometric Polynomial Higher Order Neural Network Group Models and Weighted Kernel Models for Financial Data Simulation and Prediction

2009 ◽  
pp. 484-503 ◽  
Author(s):  
Lei Zhang ◽  
Simeon J. Simoff ◽  
Jing Chun Zhang
Keyword(s):  
Neural Network ◽  
Trigonometric Polynomial ◽  
Real Life ◽  
Network Models ◽  
Higher Order ◽  
Financial Data ◽  
Suitable Model ◽  
Data Simulation ◽  
Neural Network Models ◽  
Weighted Kernel

This chapter introduces trigonometric polynomial higher order neural network models. In the area of financial data simulation and prediction, there is no single neural network model that could handle the wide variety of data and perform well in the real world. A way of solving this difficulty is to develop a number of new models, with different algorithms. A wider variety of models would give financial operators more chances to find a suitable model when they process their data. That was the major motivation for this chapter. The theoretical principles of these improved models are presented and demonstrated and experiments are conducted by using real-life financial data.

Method of complex copper-zinc ore typification using neural network models

2020 ◽  
Vol 5 ◽  
pp. 140-147 ◽  
Author(s):  
T.N. Aleksandrova ◽  
◽  
E.K. Ushakov ◽  
A.V. Orlova ◽  
◽  
...  
Keyword(s):  
Neural Network ◽  
Network Models ◽  
Neural Network Models ◽  
Copper Zinc ◽  
Complex Copper

Neural Network Techniques for Time Series Prediction: A Review

2019 ◽  
Vol 3 (3) ◽  
Author(s):  
Muhammad Faheem Mushtaq ◽  
Urooj Akram ◽  
Muhammad Aamir ◽  
Haseeb Ali ◽  
Muhammad Zulqarnain
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Time Series ◽  
Time Series Data ◽  
Weather Prediction ◽  
Time Series Prediction ◽  
Series Data ◽  
Prediction Problem ◽  
Neural Network Models ◽  
Physical Time

It is important to predict a time series because many problems that are related to prediction such as health prediction problem, climate change prediction problem and weather prediction problem include a time component. To solve the time series prediction problem various techniques have been developed over many years to enhance the accuracy of forecasting. This paper presents a review of the prediction of physical time series applications using the neural network models. Neural Networks (NN) have appeared as an effective tool for forecasting of time series.  Moreover, to resolve the problems related to time series data, there is a need of network with single layer trainable weights that is Higher Order Neural Network (HONN) which can perform nonlinearity mapping of input-output. So, the developers are focusing on HONN that has been recently considered to develop the input representation spaces broadly. The HONN model has the ability of functional mapping which determined through some time series problems and it shows the more benefits as compared to conventional Artificial Neural Networks (ANN). The goal of this research is to present the reader awareness about HONN for physical time series prediction, to highlight some benefits and challenges using HONN.

Digital twin of equipment as a basis for the consumer in digital production

2020 ◽  
Keyword(s):  
Neural Network ◽  
Tool Wear ◽  
Chip Formation ◽  
Network Models ◽  
Machining Accuracy ◽  
Neural Network Models ◽  
Digital Twin ◽  
The Neural Network ◽  
Digital Production ◽  
Cyberphysical System

The neural network models series used in the development of an aggregated digital twin of equipment as a cyber-physical system are presented. The twins of machining accuracy, chip formation and tool wear are examined in detail. On their basis, systems for stabilization of the chip formation process during cutting and diagnose of the cutting too wear are developed. Keywords cyberphysical system; neural network model of equipment; big data, digital twin of the chip formation; digital twin of the tool wear; digital twin of nanostructured coating choice

scholarly journals 4494 Predictors of Reintubation After Cardiac Surgery

2020 ◽  
Vol 4 (s1) ◽  
pp. 50-50
Author(s):  
Robert Edward Freundlich ◽  
Gen Li ◽  
Jonathan P Wanderer ◽  
Frederic T Billings ◽  
Henry Domenico ◽  
...  
Keyword(s):  
Clinical Trial ◽  
Cardiac Surgery ◽  
High Risk ◽  
Query Language ◽  
Risk Estimation ◽  
Clinical Care ◽  
Model Development ◽  
Calibration Plot ◽  
Iterative Model ◽  
Strong Candidate

OBJECTIVES/GOALS: We modeled risk of reintubation within 48 hours of cardiac surgery using variables available in the electronic health record (EHR). This model will guide recruitment for a prospective, pragmatic clinical trial entirely embedded within the EHR among those at high risk of reintubation. METHODS/STUDY POPULATION: All adult patients admitted to the cardiac intensive care unit following cardiac surgery involving thoracotomy or sternotomy were eligible for inclusion. Data were obtained from operational and analytical databases integrated into the Epic EHR, as well as institutional and departmental-derived data warehouses, using structured query language. Variables were screened for inclusion in the model based on clinical relevance, availability in the EHR as structured data, and likelihood of timely documentation during routine clinical care, in the hopes of obtaining a maximally-pragmatic model. RESULTS/ANTICIPATED RESULTS: A total of 2325 patients met inclusion criteria between November 2, 2017 and November 2, 2019. Of these patients, 68.4% were male. Median age was 63.0. The primary outcome of reintubation occurred in 112/2325 (4.8%) of patients within 48 hours and 177/2325 (7.6%) at any point in the subsequent hospital encounter. Univariate screening and iterative model development revealed numerous strong candidate predictors (ANOVA plot, figure 1), resulting in a model with acceptable calibration (calibration plot, figure 2), c = 0.666. DISCUSSION/SIGNIFICANCE OF IMPACT: Reintubation is common after cardiac surgery. Risk factors are available in the EHR. We are integrating this model into the EHR to support real-time risk estimation and to recruit and randomize high-risk patients into a clinical trial comparing post-extubation high flow nasal cannula with usual care. CONFLICT OF INTEREST DESCRIPTION: REF has received grant funding and consulting fees from Medtronic for research on inpatient monitoring.

Sign in / Sign up

Export Citation Format

Share Document