The critical need for robust decision support in the era of precision cancer therapeutics

Strengthening ‘good’ modelling practices in robust decision support: A reporting guideline for combining multiple model-based methods

Introduction to mechatronics & measurement systems

One of the major challenges for product lifecycle management systems is the lack of integrated decision support tools to help decision-making with available information in collaborative enterprise networks. Uncertainties are inherent in such networks due to lack of perfect knowledge or conflicting information. In this paper, a robust decision support approach based on imprecise probabilities is proposed. Robust Bayesian belief networks with interval probabilities are used to estimate imprecise posterior probabilities in probabilistic inference. This generic approach is demonstrated with decision-makings in design for closed-loop supply chain. The ultimate goal of robust intelligent decision support systems is to enhance the effective use of information available in collaborative engineering environments.

Background: We initiated a prospective, institution-wide study to determine whether genomic testing with a 409-gene panel in solid tumors can identify new actionable genomic information (beyond that identified by smaller hot-spot panels) and lead to enrollment in genotype matched trials using agents relevant to the alteration(s) identified when coupled with robust decision support tools. Methods: Eligible patients (pts) had no remaining standard of care therapy anticipated to extend life by more than 3 months, ECOG performance status of ≤ 1, and a willingness to consider clinical trial enrollment. The patients' tumors were initially sequenced using a hotspot panel (predominantly a 50-gene panel, Life Technology, Ion Torrent), and if no actionable alterations were found, then tumor and paired germline were sequenced with a 409-full-length (Ion Proton) gene panel. Actionable genes were defined as those for which a matched genotype selected trial exists in the institution. Results: 471 pts across more than 30 tumor types were consented and underwent 409-gene testing. Data for each mutation, relevant therapeutic agents and corresponding clinical trials were annotated. Each variant was annotated for the level of evidence that associated a specific alteration in a potentially actionable cancer gene with a potential therapeutic opportunity with appropriate references. Specific mutations, copy number variants and fusions were linked to targeted agents, clinical trials, and functional data. Data were distributed via a publicly accessible website, reports and proactive clinical trial alert notifications. Alterations in a potentially actionable gene were found in 48.0% of patients. Novel alterations in an actionable gene not found on a previous hot-spot panel were found in 36.9% of pts (174 pts). Of the 434 mutations found in actionable genes in these 174 pts, the specific variant in the gene was of known activity based on existing literature in only 17%; for 41% the variant was of unknown significance. Approximately one-quarter of patients with mutations in actionable genes were enrolled on clinical trials using matched-therapies during the period of data review. Reasons for non-enrollment were the treating physician's opinion that there was insufficient evidence for the functional significance of the variant, exclusion criteria or lack of available slots, or other reasons including pt choice. Conclusions: A significant population of patients with variants in potentially actionable cancer genes not evaluated in a traditional hot-spot cancer gene panel can be identified using a 409-gene targeted gene panel. The high number of variants of unknown significance represents a knowledge gap of clinical importance. While a number of factors contribute to bottlenecks in utilizing the expanded sequencing results, expanded genomic testing combined with robust decision support can facilitate trial enrollment. Citation Format: Kenna R. Shaw, Scott Kopetz, Vijaykumar Holla, Beate C. Litzenburger, Walter Kinyua, Blessy Sajan, J. Jack Lee, Russell Broaddus. Prospective evaluation of two-phase NGS platform coupled to active precision oncology decision support in the therapeutic management of patients with advanced cancers. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr A76.

In intelligent decision support, the integration of fuzzy logic systems with data classification has emerged as a promising avenue. This cooperative approach seeks to enhance decision-making processes by leveraging the strengths of both fuzzy logic and data classification techniques. However, a critical gap exists in the literature concerning the seamless integration of fuzzy logic systems and data classification for effective decision support. Existing approaches often treat these methodologies in isolation, overlooking the synergies that can arise from their collaborative utilization. Bridging this gap is essential for developing robust decision support systems capable of handling the intricacies of modern datasets. The research aims to address this gap by proposing a cooperative approach that seamlessly integrates fuzzy logic systems and data classification methods. By doing so, it seeks to overcome the limitations of traditional decision support systems and enhance their adaptability to real-world scenarios characterized by uncertainty and complexity. The method involves the development of a hybrid system that combines fuzzy logic rules and data classification algorithms. The fuzzy logic component captures and processes imprecise information, while the data classification component identifies patterns and trends within the data. The cooperative nature of the approach ensures that each method complements the other, resulting in a more robust and effective decision support system. The results demonstrate the improved performance of the proposed cooperative approach compared to traditional decision support systems. The system exhibits enhanced accuracy and adaptability, showcasing its potential to address the challenges posed by modern datasets.

Integrated design of a sustainable waste management system with co-modal transportation network: A robust bi-level decision support system

The use of information and models is key to making decisions related to water management, considering the interaction between the natural supply and the economic and socio-cultural systems. However this data-based decision-making is generally complex due to the uncertainties associated with these models, the various individual interests that stakeholders have regarding the water that prevail over the collective interest, and the institutional framework that frames the decisions. In a system limited by the quantity and quality of water available, and where users want to respond to their growing water needs, it requires tools that allow objective decisions to be made based on the common benefit of concurrent users. In this context, a methodological guide has been developed for the development of water resource planning processes based on data and models, which integrates the robust decision support framework (Purkey, David et al., 2018) with the development of serious games or scenathons, called guidelines for scenathons. Robust Decision Support is a framework that guides water resource planning processes through a series of steps starting from defining decision space, mapping key actors, problem formulation, tool construction, scenario definition, system vulnerability, options analysis, results exploration, and decision-making. The process includes three workshops for problem definition and vulnerability These participatory processes have shown the usefulness of having systematized information and models that make it possible, on the one hand, to understand the vulnerabilities of the system in its current condition, and to simulate scenarios of analysis of the impacts that may be generated by climate change, population growth and economic activities. In these processes, it has been understood that it is not only important to have models that accurately and precisely describe reality, but it is also fundamental how the model is built, using information and models that have credibility in the region and validating the results with the actors knowledgeable about their environment. However, interaction with stakeholders directly using the models is not easy due to multiple user profiles and knowledge. To this end, methodologies have been developed that allow interaction with complex data through visualization platforms for model results and various simulated scenarios. This interaction has been complemented with the use of serious games to generate an exchange with users using a narrative that allows transcending from existing roles and conflicts to a more purposeful dialogue. Examples of the serious games and visualization tools will be provided in the presentation https://latinoamericasei.shinyapps.io/Juego_Serio_POMCA_Campoalegre/ In this context, in the framework of the TRANSCEND project (Transformational and Robust AdaptatioN to water Scarcity and ClimatE chaNge under Deep uncertainty) we proposed a guideline for scenathons, which integrates the process of participation in the four years of the project and the development of each Scenathon for the co-creation of TAPs. The guidelines for scenathons is the roadmap to guide the process of co-creation of TAPs using models in 7 living labs and considering the associated uncertainty that may affect decision-making. We are currently developing the first year of the project where the main problems have been identified.

The growing understanding of the increased frequency and severity of extreme weather events due to climate change demands action. Locally, measures to adapt must be taken without knowing exactly what will happen, where it will happen or what the consequences will be. To meet this need, a number of decision support tools have been developed and this article investigates how municipalities can implement Robust Decision support in their urban planning. Interviews with respondents from the municipalities were conducted. After this a series of workshops were held, where an RDM method was used on local situations and follow-up interviews assessed the success and potential of the tool. Results suggest that the process addresses uncertainty, encourages bottom-up approaches and provides a tool for creating adaptive pathways in a clear and concise manner. Despite these promising findings, the success of implementation on a broader scale is seen as limited due to organizational factors.

Description of the subject. The implementation of sensor based decision support in commercial dairy herds is highly dependent on having reliable systems. Problems with sensors give missing and noisy data hampering their use. Also, the presentation of results needs to be in a form which is simple and useful. These issues were addressed using a mastitis sensor and decision support as example. Objectives. This study aims at providing and evaluating a modular system applicable to the pipeline from sensor to decision support. Method. The case of mastitis was chosen as it is of economic importance and also affects welfare of cows, and because we have worked with a commercial sensor. The problems with sensors causing missing data and noise are described and a range of filtering and monitoring modules are shown to be important to make systems functional for herd management purposes. On top of this a solid method needs to be used to interpret and present data to end users, in terms of easy to read categories. Results. Filtering and pre-adjustments of raw data are important in making algorithms robust and reliable for daily use. Re-definition of traits is needed going from traditional few groups to continuous definitions, and then to new action oriented health classes. Also, for this case focusing on mastitis, assignment to “permanently sick” groups can be helpful in keeping focus on new acute cases. Conclusions. The combined use of filtering, fix-up routines and time series models leading into action oriented categories is needed to provide simple and robust decision support. The systems may be vastly improved by opening for transmission of data between user groups and to common databases – also with a few to use data in genetic selection.

Consequential life cycle assessment (CLCA) studies how a system responds to a decision in question. There has been a growing body of CLCA studies in the last decade, with different models being incorporated from other fields, partly to compensate for the limitations of the conventional linear models used in LCA. As much as we welcome the use of new models in (C)LCA, here we provide a cautionary note on this trend by highlighting the restrictiveness of assumptions underpinning different models. And we point to a path forward for future CLCA studies. We review the model setup of, and major assumptions behind, two classes of models used in CLCA studies. One is linear models such as process- or input-output-based LCA, which have been conventionally used in LCA. And the other is nonlinear optimization models such as computable general equilibrium (CGE), which are increasingly being applied in CLCA studies. While the linear models rest on several assumptions such as fixed coefficients and unlimited supply of inputs, so do the nonlinear optimization models. Among others, CGE models assume rationality, the limitations of which have been increasingly revealed by findings of experimental and behavioral economics. We also discuss some of the foundational questions. Are LCA estimates verifiable or falsifiable? If not, then is LCA science? And is the traditional definition of science based on falsifiability suited for LCA and other disciplines studying complex systems? Considering that (1) LCA studies the complex human environment system and model estimates or predictions are largely unverifiable and (2) different classes of models have different strengths and limitations, we make the following recommendations. For decision makers, particularly policy makers, we recommend evaluating estimates from different classes of models, as opposed to relying on a single class, for more robust decision support. Each model estimate or prediction can be taken as a point of evidence. If most estimates point to the same direction, the results would be considered strong evidence of what would happen. If, on the other hand, model estimates are scattered with no obvious patterns, the results would be considered inconclusive and thus more research is needed. For modelers, we recommend efforts be put into improving a model’s predictive capability by, e.g., relaxing some of the unrealistic assumptions such as fixed input/output coefficients, 1:1 perfect displacement, and systemic optimization. Our main message is that mathematical sophistication does not necessarily equal improvement in model accuracy. Given the complexity of the human - environment system, the uncertainties of predicting the future, and the limitations of different models, a multi-model approach is entailed for more robust decision-making, and continuous effort is needed to improve model predictability.

Engineering design is essentially a collaborative decision-making process that requires rigorous evaluation, comparison and selection of design alternatives and optimization from a global perspective on the basis of different classes of design criteria. Increasing design knowledge and supporting designers to make right and intelligent decisions can achieve the improvement of the design and design efficiency. This paper develops a knowledge-based decision support model and framework that can be extensively applied for an engineering system, which allows for the seamless / smooth integration of collaborative product development with optimal product performance. The developed hybrid robust design decision support model quantitatively incorporates qualitative design knowledge and preferences of multiple, conflicting attributes stored in a knowledge repository so that a better understanding of the consequences of design decisions can be achieved from an overall perspective. Two new concepts and mechanisms, transforming bridge and regulatory switch, are introduced in integration of decision support models. The results of this work provide a framework for an efficient decision support environment involving distributed resources to shorten the realization of products with optimal life-cycle performance and competitiveness. The developed methodology and framework are generic and flexible enough to be used in a variety of decision problems. Case application and studies for concept evaluation and selection in design for mass customization are provided.

Forecasting the NYSE composite index with technical analysis, pattern recognizer, neural network, and genetic algorithm: a case study in romantic decision support

Antibody-Drug Conjugates (ADCs) for Breast Cancer Therapeutic Landscape: Concept and Mechanisms of Action.

Handling uncertainties is a major challenge in climate change adaptation. A variety of robust decision support approaches that aim for better management of uncertainty have recently been emerging and are used in environmental planning. The present study examined to what extent existing processes of planning for future sea-level rise in Sweden utilised similar approaches. Three core principles of robust decision support approaches were identified and used as a tool for analyzing five cases of planning for future sea-level rise in companies and authorities at different levels in society. The results show that planning processes typically do not embrace uncertainties, do not use a bottom-up approach and do not specifically aim for robustness, which points to a discrepancy between current planning paradigms and the core principles of robust decision support approaches.

In this research, we utilize semantic technology for robust early diagnosis and decision support. We present a light-weight platform that provides the enduser with direct access to the data through an ontology, and enables detection of any forthcoming faults by considering the data only from the reliable sensors. Concurrently, it indicates the actual sources of the detected faults, enabling mitigation action to be taken. Our work is focused on systems that require only real-time data and a restricted part of the historic data, such as fuel cell stack systems. First, we present an upper-level ontology that captures the semantics of such monitored systems and then we present the structure of the platform. Next, we specialize on the fuel cell paradigm and we provide a detailed description of our platform’s functionality that can aid future servicing problem reporting applications.