Network Decision Support Research Articles

Assessing Serious Spinal Pathology Using Bayesian Network Decision Support: Development and Validation Study.

Identifying and managing serious spinal pathology (SSP) such as cauda equina syndrome or spinal infection in patients presenting with low back pain is challenging. Traditional red flag questioning is increasingly criticized, and previous studies show that many clinicians lack confidence in managing patients presenting with red flags. Improving decision-making and reducing the variability of care for these patients is a key priority for clinicians and researchers. We aimed to improve SSP identification by constructing and validating a decision support tool using a Bayesian network (BN), which is an artificial intelligence technique that combines current evidence and expert knowledge. A modified RAND appropriateness procedure was undertaken with 16 experts over 3 rounds, designed to elicit the variables, structure, and conditional probabilities necessary to build a causal BN. The BN predicts the likelihood of a patient with a particular presentation having an SSP. The second part of this study used an established framework to direct a 4-part validation that included comparison of the BN with consensus statements, practice guidelines, and recent research. Clinical cases were entered into the model and the results were compared with clinical judgment from spinal experts who were not involved in the elicitation. Receiver operating characteristic curves were plotted and area under the curve were calculated for accuracy statistics. The RAND appropriateness procedure elicited a model including 38 variables in 3 domains: risk factors (10 variables), signs and symptoms (17 variables), and judgment factors (11 variables). Clear consensus was found in the risk factors and signs and symptoms for SSP conditions. The 4-part BN validation demonstrated good performance overall and identified areas for further development. Comparison with available clinical literature showed good overall agreement but suggested certain improvements required to, for example, 2 of the 11 judgment factors. Case analysis showed that cauda equina syndrome, space-occupying lesion/cancer, and inflammatory condition identification performed well across the validation domains. Fracture identification performed less well, but the reasons for the erroneous results are well understood. A review of the content by independent spinal experts backed up the issues with the fracture node, but the BN was otherwise deemed acceptable. The RAND appropriateness procedure and validation framework were successfully implemented to develop the BN for SSP. In comparison with other expert-elicited BN studies, this work goes a step further in validating the output before attempting implementation. Using a framework for model validation, the BN showed encouraging validity and has provided avenues for further developing the outputs that demonstrated poor accuracy. This study provides the vital first step of improving our ability to predict outcomes in low back pain by first considering the problem of SSP. RR2-10.2196/21804.

Optimizing Textile Manufacturing With Neural Network Decision Support: An Ornstein-Uhlenbeck Reinforcement Learning Approach

The textile sector makes a substantial contribution to the global economy. During the Industry 4.0 era, the textile production process is anticipated to become more adaptable and versatile. The complex interplay between the macroscopic factors in different textile processes may make decision-making challenging. To address these challenges and provide a viable answer to textile manufacturing companies, machine learning methods are suggested to develop decision support systems. It is possible for a model to correctly reflect the intricate relationships that exist between the parameters and performances of a textile manufacturing process. This representation serves as the foundation for a decision support system. Especially in this work, quality control can be improved by predicting the defects in fabrics using an effective Ornstein-Uhlenbeck reinforcement learning-based decision neural network (OU_RL_DNN), which can detect both local and global defects. The framework's approach effectively reduces the expense of manually annotating the dataset. It takes a few faulty samples mixed with reference samples to learn fault characteristics and properly locate them. The experiment, which comprised a data set that was available to the public and three private fabric data sets, demonstrated that the proposed method surpasses a number of cutting-edge methods in terms of detection and detection time.

The implementation of NILS: A web-based artificial neural network decision support tool for noninvasive lymph node staging in breast cancer.

To implement artificial neural network (ANN) algorithms for noninvasive lymph node staging (NILS) to a decision support tool and facilitate the option to omit surgical axillary staging in breast cancer patients with low-risk of nodal metastasis. The NILS tool is a further development of an ANN prototype for the prediction of nodal status. Training and internal validation of the original algorithm included 15 clinical and tumor-related variables from a consecutive cohort of 800 breast cancer cases. The updated NILS tool included 10 top-ranked input variables from the original prototype. A workflow with four ANN pathways was additionally developed to allow different combinations of missing preoperative input values. Predictive performances were assessed by area under the receiver operating characteristics curves (AUC) and sensitivity/specificity values at defined cut-points. Clinical utility was presented by estimating possible sentinel lymph node biopsy (SLNB) reduction rates. The principles of user-centered design were applied to develop an interactive web-interface to predict the patient's probability of healthy lymph nodes. A technical validation of the interface was performed using data from 100 test patients selected to cover all combinations of missing histopathological input values. ANN algorithms for the prediction of nodal status have been implemented into the web-based NILS tool for personalized, noninvasive nodal staging in breast cancer. The estimated probability of healthy lymph nodes using the interface showed a complete concordance with estimations from the reference algorithm except in two cases that had been wrongly included (ineligible for the technical validation). NILS predictive performance to distinguish node-negative from node-positive disease, also with missing values, displayed AUC ranged from 0.718 (95% CI, 0.687-0.748) to 0.735 (95% CI, 0.704-0.764), with good calibration. Sensitivity 90% and specificity 34% were demonstrated. The potential to abstain from axillary surgery was observed in 26% of patients using the NILS tool, acknowledging a false negative rate of 10%, which is clinically accepted for the standard SLNB technique. The implementation of NILS into a web-interface are expected to provide the health care with decision support and facilitate preoperative identification of patients who could be good candidates to avoid unnecessary surgical axillary staging.

Application of a Multi-Layer Perceptron in Preoperative Screening for Orthognathic Surgery.

Orthognathic surgery is used to treat moderate to severe occlusal discrepancies. Examinations and measurements for preoperative screening are essential procedures. A careful analysis is needed to decide whether cases require orthognathic surgery. This study developed screening software using a multi-layer perceptron to determine whether orthognathic surgery is required. In total, 538 digital lateral cephalometric radiographs were retrospectively collected from a hospital data system. The input data consisted of seven cephalometric variables. All cephalograms were analyzed by the Detectron2 detection and segmentation algorithms. A keypoint region-based convolutional neural network (R-CNN) was used for object detection, and an artificial neural network (ANN) was used for classification. This novel neural network decision support system was created and validated using Keras software. The output data are shown as a number from 0 to 1, with cases requiring orthognathic surgery being indicated by a number approaching 1. The screening software demonstrated a diagnostic agreement of 96.3% with specialists regarding the requirement for orthognathic surgery. A confusion matrix showed that only 2 out of 54 cases were misdiagnosed (accuracy = 0.963, sensitivity = 1, precision = 0.93, F-value = 0.963, area under the curve = 0.96). Orthognathic surgery screening with a keypoint R-CNN for object detection and an ANN for classification showed 96.3% diagnostic agreement in this study.

A Deep Multi-Label Segmentation Network For Eosinophilic Esophagitis Whole Slide Biopsy Diagnostics.

Eosinophilic esophagitis (EoE) is an allergic inflammatory condition of the esophagus associated with elevated numbers of eosinophils. Disease diagnosis and monitoring require determining the concentration of eosinophils in esophageal biopsies, a time-consuming, tedious and somewhat subjective task currently performed by pathologists. Here, we developed a machine learning pipeline to identify, quantitate and diagnose EoE patients' at the whole slide image level. We propose a platform that combines multi-label segmentation deep network decision support system with dynamics convolution that is able to process whole biopsy slide. Our network is able to segment both intact and not-intact eosinophils with a mean intersection over union (mIoU) of 0.93. This segmentation enables the local quantification of intact eosinophils with a mean absolute error of 0.611 eosinophils. We examined a cohort of 1066 whole slide images from 400 patients derived from multiple institutions. Using this set, our model achieved a global accuracy of 94.75%, sensitivity of 94.13%, and specificity of 95.25% in reporting EoE disease activity. Our work provides state-of-the-art performances on the largest EoE cohort to date, and successfully addresses two of the main challenges in EoE diagnostics and digital pathology, the need to detect several types of small features simultaneously, and the ability to analyze whole slides efficiently. Our results pave the way for an automated diagnosis of EoE and can be utilized for other conditions with similar challenges.

Evaluating the Optimal External Equity Financing Strategy and Critical Factors for the Startup of Lending Company in Taiwan: An Application of Expert Network Decision Model

During enterprise foundation and development, internal finance and debt finance are of vital importance to start-up entrepreneurs. Therefore, the purpose of this study is mainly to focus on how start-ups can make the optimal evaluation among different external equity crowdfunding solutions and to establish a network decision support model that evaluates the optimal financing solution of start-ups for external equity crowdfunding based on decision science and network architecture. The Lending Company in Financial Technology Industry (LCFTI) was taken as an example. The results indicate that equity crowdfunding is the optimal financing plan in LCFTI. Academically, the results of this study not only help propose a network decision support model using decision science methods and implementing the network analysis to establish an architecture to evaluate the optimal financing plans of start-ups for external equity crowdfunding, they also makes up for the gap in the optimal financing plans of entrepreneurs or start-ups for external equity financing, which has not been specified in the POT theory in the past. Practically, this study provides a useful tool for the entrepreneur of LCFTI to understand the key factors affecting the optimal financing plans for external equity financing and enables LCFTI to measure the optimal financing plans for external equity financing to improve the success rate of finance.

SENSITIVITY ANALYSIS OF SUPPLY CHAIN MANAGEMENT PROBLEM BASED ON HIERARCHICAL AND NETWORK DECISION SUPPORT MODELS

Background. Predicting future sales is necessary to control the flow of goods in supply chains (SC), which is why firms perform forecasting of consumer demand. To study the reliability of the solution, the sensitivity analysis (SA) of the results to changes in the original data is performed. Network decision support (DS) models usually contain a very large number of elements and connections between them, which makes it difficult to perform SA. Objective. The aim of the paper is to assess the sensitivity of the solution, given in the form of the element ranking of the network DS model, to inaccuracies and contradictions in elements of expert pairwise comparison matrices (PCMs), as well as to changes in individual elements of the supermatrix of the network DS model. To assess the priority of different types of information in the SC management system and to assess the sensitivity of the decision to obtain a more accurate forecast of consumer demand. Methods. Evaluation of DS models to prioritize information needs in the SC management system is carried out based on a developed method of network analysis. The method for assessing the sensitivity of the decision, proposed for the hierarchical DS model, includes finding stable elements of each level of the hierarchy and assessing the degree of sensitivity of global ranking of elements. The sensitivity of a solution based on a network DS model includes an assessment of the stability of local rankings, finding the matrix elements that most affect the change in consistency and change in the local ranking, as well as the sensitivity to changes in individual elements of the model supermatrix. Results. The method of complex assessment of sensitivity has been further developed, the stages of assessing the stability of local ranking of elements of the network DS model and the resilience of the pairwise comparison matrix elements to changes in the permissible inconsistency have been improved. A method for SA of results based on the network DS model to changes in individual elements of the supermatrix using machine learning tools is generalized. Conclusions. In the SC management problem, the elements of the expert PCM are found, which to the greatest extent affect the change in consistency and change in the rankings of decision alternatives. The stable elements of the matrix and elements for revision by an expert were calculated in order to increase the reliability of the solution to the SC management problem based on the hierarchical and network DS models.

A Bayesian Network Decision Support Tool for Low Back Pain Using a RAND Appropriateness Procedure: Proposal and Internal Pilot Study.

BackgroundLow back pain (LBP) is an increasingly burdensome condition for patients and health professionals alike, with consistent demonstration of increasing persistent pain and disability. Previous decision support tools for LBP management have focused on a subset of factors owing to time constraints and ease of use for the clinician. With the explosion of interest in machine learning tools and the commitment from Western governments to introduce this technology, there are opportunities to develop intelligent decision support tools. We will do this for LBP using a Bayesian network, which will entail constructing a clinical reasoning model elicited from experts.ObjectiveThis paper proposes a method for conducting a modified RAND appropriateness procedure to elicit the knowledge required to construct a Bayesian network from a group of domain experts in LBP, and reports the lessons learned from the internal pilot of the procedure.MethodsWe propose to recruit expert clinicians with a special interest in LBP from across a range of medical specialties, such as orthopedics, rheumatology, and sports medicine. The procedure will consist of four stages. Stage 1 is an online elicitation of variables to be considered by the model, followed by a face-to-face workshop. Stage 2 is an online elicitation of the structure of the model, followed by a face-to-face workshop. Stage 3 consists of an online phase to elicit probabilities to populate the Bayesian network. Stage 4 is a rudimentary validation of the Bayesian network.ResultsEthical approval has been obtained from the Research Ethics Committee at Queen Mary University of London. An internal pilot of the procedure has been run with clinical colleagues from the research team. This showed that an alternating process of three remote activities and two in-person meetings was required to complete the elicitation without overburdening participants. Lessons learned have included the need for a bespoke online elicitation tool to run between face-to-face meetings and for careful operational definition of descriptive terms, even if widely clinically used. Further, tools are required to remotely deliver training about self-identification of various forms of cognitive bias and explain the underlying principles of a Bayesian network. The use of the internal pilot was recognized as being a methodological necessity.ConclusionsWe have proposed a method to construct Bayesian networks that are representative of expert clinical reasoning for a musculoskeletal condition in this case. We have tested the method with an internal pilot to refine the process prior to deployment, which indicates the process can be successful. The internal pilot has also revealed the software support requirements for the elicitation process to model clinical reasoning for a range of conditions.International Registered Report Identifier (IRRID)DERR1-10.2196/21804

Reliability Based Life Cycle Management of Bridge Subjected to Fatigue Damage

In recent times, there has been an increase in transport infrastructure failure. This increase is due to aging infrastructure, increased number of extreme weather events caused by climate change, and increased traffic loading. Accordingly, the need for planned and unplanned maintenance interventions is rising. Associated costs do not only involve direct maintenance or reconstruction costs, but also secondary effects experienced by users of the transport network as well as the environment and society in general. Infrastructure managers require tools for accurate quantification of infrastructure resilience that will enable rational adaptation investment strategies, so as to maintain high level of safety of transport networks. Through the development of a Global Safety Framework, at the core of which is a Multi-modal Network Decision Support Tool, the SAFE-10-T project (Safety of Transport Infrastructure on the TEN-T Network) is providing integrated solutions to issues related to infrastructure safety and planning. The paper presents a reliability-based whole life cycle model developed within this project enabling strategic investment decisions that maximize safety, minimize disruption, and environmental impacts and allow for the best use of limited resources. The model is applied on a case study of a bridge in the Port of Rotterdam in the Netherlands.

National Comprehensive Cancer Network (NCCN) Decision Support Tool for Patients With NSCLC

National Comprehensive Cancer Network (NCCN) Decision Support Tool for Patients With NSCLC

Aspects of fuzzy-neural network decision support in identifying the author of handwriting

Aspects of fuzzy-neural network decision support in identifying the author of handwriting

Toward an integrated Bayesian network approach to measurement error detection and correction

ABSTRACTIn this article, the quality of data produced by national statistical institutes and by governmental institutions is considered. In particular, the problem of measurement error is analyzed and an integrated Bayesian network decision support system based on non-parametric Bayesian networks is proposed for its detection and correction. Non-parametric Bayesian networks are graphical models expressing dependence structure via bivariate copulas associated to the edges of the graph. The network structure and the misreport probability are estimated using a validation sample. The Bayesian network model is proposed to decide: (i) which records have to be corrected; (ii) the kind and amount of correction to be adopted. The proposed correction procedure is applied to the Banca d’Italia Survey on Household Income and Wealth and, specifically, the bond amounts are analyzed. Finally, the sensitivity of the conditional distribution of the true value random variable given the observed one to different evidence configurations is studied.

ON CONSTRUCTION OF AN INTELLIGENT SUBSYSTEM FOR ANALYZING THE PARAMETERS OF A MARSHALLING HUB

For the English abstract and full text of the article please see the attached PDF-File (English version follows Russian version).The work was supported by the Russian Fundamental Research Foundation, project 17-20-01133 ofi_m_ Russian Railways. ABSTRACT The article considers the issues of continuous monitoring of the situation at the marshalling yard and detection of the possibility of occurrence of dangerous situations. An approach is proposed for constructing an automated intellectual subsystem for analyzing and timely forecasting the critical utilization of railway sorting units. The solution of the problem is proposed with the help of network communication technologies due to the use of information from automated data collection systems and the neural network decision support subsystem. Keywords: transport, intellectual system, system analysis, forecasting of critical situations, simulation modeling, sorting node, neural network, information, communication technologies.

Artificial neural network decision support tool for assessment of the energy performance and the refurbishment actions for the non-residential building stock in Southern Italy

The public buildings sector represents one of the most intensive items of EU energy consumption; the application of retrofit solutions in existing buildings is a crucial way to reduce its impact. To facilitate the knowledge of the energy performance of existing non-residential buildings and the choice of the more adequate actions, Public Administrations (PA) should have the availability of proper tools. Within the Italian project “POI 2007-13”, a database and a decision support tool, for easy use, even to a non-technical user, have been developed. A large set of data, obtained from the energy audits of 151 existing public buildings located in four regions of South Italy have been analysed, elaborated, and organised in a database. This was used to identify the best architectures of two ANNs and to train them. The first ANN provides the actual energy performance of any building; the second ANN assesses key economic indicators. A decision support tool, based on the use of these ANNs is conceived for a fast prediction of the energy performance of buildings and for a first selection of energy retrofit actions that can be applied.

Using automatically extracted information from mammography reports for decision-support

ObjectiveTo evaluate a system we developed that connects natural language processing (NLP) for information extraction from narrative text mammography reports with a Bayesian network for decision-support about breast cancer diagnosis. The ultimate goal of this system is to provide decision support as part of the workflow of producing the radiology report. Materials and methodsWe built a system that uses an NLP information extraction system (which extract BI-RADS descriptors and clinical information from mammography reports) to provide the necessary inputs to a Bayesian network (BN) decision support system (DSS) that estimates lesion malignancy from BI-RADS descriptors. We used this integrated system to predict diagnosis of breast cancer from radiology text reports and evaluated it with a reference standard of 300 mammography reports. We collected two different outputs from the DSS: (1) the probability of malignancy and (2) the BI-RADS final assessment category. Since NLP may produce imperfect inputs to the DSS, we compared the difference between using perfect (“reference standard”) structured inputs to the DSS (“RS-DSS”) vs NLP-derived inputs (“NLP-DSS”) on the output of the DSS using the concordance correlation coefficient. We measured the classification accuracy of the BI-RADS final assessment category when using NLP-DSS, compared with the ground truth category established by the radiologist. ResultsThe NLP-DSS and RS-DSS had closely matched probabilities, with a mean paired difference of 0.004±0.025. The concordance correlation of these paired measures was 0.95. The accuracy of the NLP-DSS to predict the correct BI-RADS final assessment category was 97.58%. ConclusionThe accuracy of the information extracted from mammography reports using the NLP system was sufficient to provide accurate DSS results. We believe our system could ultimately reduce the variation in practice in mammography related to assessment of malignant lesions and improve management decisions.

Clinical evidence framework for Bayesian networks

There is poor uptake of prognostic decision support models by clinicians regardless of their accuracy. There is evidence that this results from doubts about the basis of the model as the evidence behind clinical models is often not clear to anyone other than their developers. In this paper, we propose a framework for representing the evidence-base of a Bayesian network (BN) decision support model. The aim of this evidence framework is to be able to present all the clinical evidence alongside the BN itself. The evidence framework is capable of presenting supporting and conflicting evidence, and evidence associated with relevant but excluded factors. It also allows the completeness of the evidence to be queried. We illustrate this framework using a BN that has been previously developed to predict acute traumatic coagulopathy, a potentially fatal disorder of blood clotting, at early stages of trauma care.

A Neural Network Decision Support System for Analysing Markets

Today, markets are equipped with IT-based systems to facilitate the flow of information within markets and to provide useful information for producers and costumers. Therefore, real time decision making is a significant issue of IT environment for obtaining maximum profit in markets. A valuable tool for real time decision making are Decision Support Systems (DSSs). Here, we propose a DSS to identify a set of optimal markets for a producer. The producer aims to determine the markets that provide more profit for him via information systems of markets that analyze all transactions and prepare reports. Due to these reports the producer would decide about markets that provide the maximum profit. The effectiveness of the proposed integrated model is illustrated through numerical example.

Visual analysis of supply network risks: Insights from the electronics industry

In today's complex, global supply networks it has become increasingly challenging to identify, evaluate, and mitigate risks of disruption. Traditional supply chain practices have primarily focused on dyadic risk management, rarely considering risks in the sub-tier supply network. However, this approach severely limits a decision maker's ability to understand the highly interconnected nature of systemic risks and develop corresponding mitigation strategies. Grounded in theories of supply chains as complex systems, network analysis, and risk management, we demonstrate the importance of visual decision support for supply network risk assessment. We empirically illustrate our approach with supply network visualization examples from the electronics industry. We conclude the study with implications for the design and implementation of visual supply network decision support systems and future research opportunities.

Real‐time modeling of water distribution systems: A case study

Water utilities worldwide face increasing challenges to preserve the hydraulic and water quality integrity of their water distribution networks. These challenges stem from burgeoning populations and migration to urban cities that continue to increase the load on aging, inefficient, and already strained infrastructures. This has created a pressing need for integrating supervisory control and acquisition systems with network simulation models for proactive management of these networks. Such an integrated platform is the basis for the real‐time smart water network decision support system (SWNDSS) described here. The proposed system has the power to transform a water utility's routine network modeling functions from planning and design to full‐spectrum engagement that drives more efficient operations—including managing water quality and energy, developing daily operating plans, addressing planned and emergency outages, and diagnosing and resolving field issues. Aspects of the SWNDSS ar described in a case study from the Las Vegas Valley Water District in Las Vegas, Nev.

A Bayesian Network Decision Support System for Order Management in New Product Development

Recently, the way firms enter to markets is influenced by the customer requirements applying product improvement. Thus, market-driven product design and development is now a favorite research topic in the literature. Order management prediction for a new product help firms to overcome the future uncertainties. Here, we propose a decision support system for customers' order management in a new product development process. The drawback of complicated decision support problems are the complexities involved in interpreting causal relationships among decision variables. Therefore, Bayesian Network (BN) has shown excellent decision support competence due to its flexible structure allowing itself to extract appropriate and robust causal relationships among target variable and related explanatory variables. We make use of a decision support BN as a prediction aid for order management in a new product development process. system predicts customer purchasing behavior using a system dynamics approach based on three pieces of information: product attractiveness, customer preferences and satisfaction, and marketing strategy. It also estimates the long-term NCLV based on Markov analysis. This can help managers to determine which product will be most lucrative to launch and the kinds of marketing strategies that should be adopted for the new product. It also helps improve new product development in the future by collating up to date information on market and product attributes. In recent years, many conventional and market-based decision support systems for product design have been developed (3-7). New product development, from idea creation to product introduction, requires inter-departmental communication among designers, engineers, and marketing personnel. These highlight the key areas that ought to be considered in making decisions on new product development, including customer requirements, customer satisfaction, market demand, product quality, product design, and pricing. However, no decision support system takes all of the key areas into account at the same time. Furthermore, to achieve a competitive edge in a market, sensible decisions must be made about various aspects of new product development, such as product attributes, customer segment, and promotion and marketing strategies.