Decision Support Model Research Articles

Sales strategies with a probabilistic business model of an insurance company: The role of updating and targeting

The major goal of decisions made by a business organization is to enhance business performance. These days, owners, managers and other stakeholders are seeking for opportunities of modelling and automating decisions by analysing the most recent data with the help of artificial intelligence (AI). This study outlines a simple theoretical model framework using internal and external information on current and potential clients and performing calculations followed by immediate updating of contracting probabilities after each sales attempt. This can help increase sales efficiency, revenues, and profits in an easily programmable way and serve as a basis for focusing on the most promising deals customising personal offers of best-selling products for each potential client. The search for new customers is supported by the continuous and systematic collection and analysis of external and internal statistical data, organising them into a unified database, and using a decision support model based on it. As an illustration, the paper presents a fictitious model setup and simulations for an insurance company considering different regions, age groups and genders of clients when analysing probabilities of contracting, average sales and profits per contract. The elements of the model, however, can be generalised or adjusted to any sector. Results show that dynamic targeting strategies based on model calculations and most current information outperform static or non-targeted actions. The process from data to decision-making to improve business performance and the decision itself can be easily algorithmised. The feedback of the results into the model carries the potential for automated self-learning and self-correction. The proposed framework can serve as a basis for a self-sustaining artificial business intelligence system.

COMBINED DECISION SUPPORT MODEL FOR DEVELOPMENT PRIORITIES OF THE KYIV CAR TUNNEL NETWORK

An analysis is conducted for the opportunities to reduce climate (ecological), technogenic and safety risks in the Kyiv transport infrastructure in case of construction of the car tunnel system envisioned in the Kyiv city General plan. The involved structural and functional factors included district type in the potential tunnel’s area, downtown factor, population density, daily pendular movement, arterial network density, aver-age traffic speed at peak hours, surface connectivity of the exits (portals) of the tunnel by automobile roads, surface road throughput, criticalness of the transport infrastructure object, and vulnerability in case of military attacks or terrorist acts. The modified morphological analysis method was applied in the research, that con-sidered groups of climate (ecological) and techogenic risks, and also the military (terrorist) threats. This be-came the basis for construction of the decision support model regarding the evaluation of priorities and rank-ings (rational order) of tunnel construction. Additionally, a combination of morphological analysis with analyt-ical hierarchy process was proposed that provided alternative viewpoints with a different ranking compared to the basic model, depending on the given priorities (weights) of risk factors during decision making. This approach may be useful for development of the transport policy with the goal of approaching climate neutrality for big cities, as well as maximum protection of urban transport infrastructure in conditions of wartime threats.Using the developed system tools and combined models, the urban society can become safer and more resilient, managing transport policies on the basis of the system approach, and multi-criteria analysis of im-pact factors.

Deep learning models for MRI-based clinical decision support in cervical spine degenerative diseases

PurposeThe purpose of our study is to develop a deep learning (DL) model based on MRI and analyze its consistency with the treatment recommendations for degenerative cervical spine disorders provided by the spine surgeons at our hospital.MethodsIn this study, MRI of patients who were hospitalized for cervical spine degenerative disorders at our hospital from July 2023 to July 2024 were primarily collected. The dataset was divided into a training set, a validation set, and an external validation set. Four versions of the DL model were constructed. The external validation set was used to assess the consistency between the DL model and spine surgeons’ recommendations about indication of cervical spine surgery regarding the dataset.ResultsThis study collected a total of 756 MR images from 189 patients. The external validation set included 30 patients and a total of 120 MR images, consisting of 43 images for grade 0, 20 images for grade 1, and 57 images for grade 2. The region of interest (ROI) detection model completed the ROI detection task perfectly. For the binary classification (grades 0 and 1, 2), DL version 1 showed the best consistency with the spine surgeons, achieving a Cohen’s Kappa value of 0.874. DL version 4 also achieved nearly perfect consistency, with a Cohen’s Kappa value of 0.811. For the three-class classification, DL version 1 demonstrated the best consistency with the spine surgeons, achieving a Cohen’s Kappa value of 0.743, while DL version 2 and DL version 4 also showed substantial consistency, with Cohen’s Kappa values of 0.615 and 0.664, respectively.ConclusionWe initially developed deep learning algorithms that can provide clinical recommendations based on cervical spine MRI. The algorithm shows substantial consistency with experienced spine surgeons.

Real-time decision support model for logistics of emergency patient transfers from hospitals via an integrated optimisation and machine learning approach

Real-time decision support model for logistics of emergency patient transfers from hospitals via an integrated optimisation and machine learning approach

Systematic review of modelling techniques in carbon trading research in construction

Purpose Carbon emissions trading is an effective instrument in reducing greenhouse gas emissions. There is a notable scarcity of comprehensive reviews on the modelling techniques within carbon trading research in construction. Design/methodology/approach This paper reviews 68 relevant articles published in 19 peer-reviewed journals using systematic search. Scientometric analysis and content analysis are undertaken. Findings Generally, China was the largest contributor to carbon trading research using quantitative models (representing 36% of the total articles). From the results, the modelling techniques identified were multi-objective grasshopper optimisation algorithm; system dynamics; interpretive structural modelling; multi-agent-based model; decision-support model; multi-objective chaotic sine cosine algorithm; optimised backpropagation neural network; sequential panel selection method; Granger causality test; and impulse response analysis. Moreover, the advantages and disadvantages of these techniques were identified. System dynamics was recommended as the most suitable modelling technique for carbon trading in construction. Originality/value This study is significant, and through this review paper, practitioners can easily be more familiar with the significant modelling techniques, and this will motivate them to better understand their uses.

Management of the Portfolio of Works For A Public Hospital Complex

Objectives: The main objective of this research is to develop a decision-making support model for prioritizing works projects in a public hospital complex, establishing criteria for characterizing and evaluating these projects. Theoretical reference: Strategic management; management control; project portfolio management; multi-criteria analysis; hospital works projects. Method: This is a quantitative exploratory research based on the Multi-Criterion Decision Support Methodology (MCDA), whose primary data are derived from semi-structured questionnaires. This methodology makes it possible to carry out a multi-criterion analysis of a large quantity of information, involving conflicting interests amongst themselves in decision-making contexts of high uncertainty and high complexity. Results and Discussion: A model for evaluating demands was developed, consisting of 16 criteria for analysis (descriptors), grouped into four thematic groups, whose measurement scales, once transformed into cardinal quantities, present the performance of each characteristic under study and make possible the construction of a global model for evaluating demands. These criteria are based on the decision-makers' perceptions of the institutional needs of works projects, revealing their concerns about the infrastructure problems in the body's buildings. Implications of Research: The implementation of this model leads to the creation of a service queue for the demands of works projects that materializes the priority actions in the area of infrastructure from the objectives, expectations and preferences of the decision makers. In the end, the model developed makes it possible, starting from the attribution and weighting of values, to define institutional priorities. Originality/Value: The research points to a path for overcoming major challenges in the decision-making on the prioritization of works projects in a public hospital, highlighting the high volume of demands, conflicts of interest, resource constraint and obsolescence of buildings and equipment.

Collaborative Scheduling of Mass Rescue Operations at Sea

The continued growth of the global cruise market has posed a major challenge to maritime search and rescue, as available rescue resources are limited compared to the scale of passenger ship disasters. To bridge the gap between the number of passengers in distress and the availability of rescue resources, this paper develops a two-stage Mass Rescue Operation (MRO) decision support model (MRO model) to fully utilize the available multiple rescue resources. Based on the combinatorial optimization theory, in the MRO model we consider the rescue capacity of multiple rescue resources and the synergy between them, the accident types and the marine environment conditions to optimize two objectives (rescue time and number of rescue resources dispatched) in two stages. In the first stage, the objective is to minimize the rescue time by Classical Selection Sort Algorithm. In the second stage, the rescue time and the number of rescue resources are simultaneously optimized by Simulated Annealing Arithmetic (SAA) integrated with Genetic Algorithm (GA). Furthermore, considering the actual role of helicopters in MROs, the MRO model is enhanced to schedule helicopters mandatorily or non-mandatorily. Finally, the MRO model was verified by simulating accidents in the Taiwan Strait. The simulation results show that compared with the first stage, the rescue time in the second stage model is saved by up to 16.18% and the number of rescue resources is reduced by up to 45.16%.

Prioritizing sustainability of renewable energy projects in Morocco

Purpose The purpose of the study is to examine the electricity generation potential of different renewable energy sources in Morocco, as well as to propose a novel decision support model to prioritize renewable energy alternatives. The aim is to provide recommendations that support Morocco’s energy transition strategy and assist policymakers in devising future sustainable energy policies. Design/methodology/approach The paper integrates the analytic hierarchy process with expert feedback to prioritize five renewable energy alternatives – solar photovoltaics (PV), concentrated solar panels (CSP), wind, hydro and biomass – specific to the Moroccan context. The model uses 14 sub-criteria clustered into four main categories: economic, technological, socio-political and environmental. The criteria were weighted based on inputs from local energy sector experts to capture the distinct priorities and contextual specificities of Morocco’s renewable energy landscape. Findings The study’s findings emphasize the dominance of economic criteria among Moroccan experts. Solar PV emerged as the most suitable option due to its cost-effectiveness and alignment with Morocco’s policies to support solar energy, followed by hydropower and wind. CSP and biomass ranked lowest due to high costs and limited viability. Research limitations/implications This study highlighted the need for quality data to support informed decision-making, as well as for a centralized database to facilitate access to consistent information on renewable energy projects. Practical implications It is recommended that policymakers prioritize mature and cost-effective technologies by establishing supportive regulatory frameworks and expanding investments in hybrid renewable energy infrastructures. Originality/value In addition to introducing the Analytic Hierarchy Process model for the first time within the Moroccan context, to the best of the authors’ knowledge, this research draws recommendations to assist Moroccan policymakers in devising future sustainable policies for the energy sector.

Resilient transportation network design with disruption uncertainty and lead times

Cost-efficient and reliable transports are needed to supply products competitively. Thus, particularly in increasingly complex and global supply chains, identifying the optimal transportation mode is a critical decision. Transportation modes, however, are prone to disruptions, such as hurricanes, low water levels, or port shutdowns, resulting in transportation stops and cost increases. To counteract these disruptions, different resilience strategies are studied to increase the capability of a network to withstand, adapt, and recover from disruptions. For a cost-optimal use, it is necessary to determine the optimal mix of strategic, tactical, and operational strategies.We provide a decision-support model that decides on the optimal mix of resilience strategies, such as multi-sourcing, inventory, or operational re-routing, for a supply chain with transportation disruption uncertainty to minimize total expected costs. The problem is formulated as a two-stage stochastic mixed-integer linear program that explicitly considers lead times. To handle large instances, we propose a Benders decomposition approach enhanced through lower-bound lifting and valid inequalities, branch-and-benders-cut, and a warm-start heuristic. Computational experiments show that large instances can be solved to near-optimality, whereas a commercial solver does not find feasible solutions.We present a case study for a company’s inbound supply chain design with recurring transportation cost uncertainty. Considering disruption and lead time effects, a mix of resilience strategies from strategic to operational level leads to cost improvements of up to 50%. Furthermore, we show that the ability to predict disruptions can further reduce resilience-related costs by 10% if sufficient operational re-routing capacities are available.

Racial differences in laboratory testing as a potential mechanism for bias in AI: A matched cohort analysis in emergency department visits.

AI models are often trained using available laboratory test results. Racial differences in laboratory testing may bias AI models for clinical decision support, amplifying existing inequities. This study aims to measure the extent of racial differences in laboratory testing in adult emergency department (ED) visits. We conducted a retrospective 1:1 exact-matched cohort study of Black and White adult patients seen in the ED, matching on age, biological sex, chief complaint, and ED triage score, using ED visits at two U.S. teaching hospitals: Michigan Medicine, Ann Arbor, MI (U-M, 2015-2022), and Beth Israel Deaconess Medical Center, Boston, MA (BIDMC, 2011-2019). Post-matching, White patients had significantly higher testing rates than Black patients for complete blood count (BIDMC difference: 1.7%, 95% CI: 1.1% to 2.4%, U-M difference: 2.0%, 95% CI: 1.6% to 2.5%), metabolic panel (BIDMC: 1.5%, 95% CI: 0.9% to 2.1%, U-M: 1.9%, 95% CI: 1.4% to 2.4%), and blood culture (BIDMC: 0.9%, 95% CI: 0.5% to 1.2%, U-M: 0.7%, 95% CI: 0.4% to 1.1%). Black patients had significantly higher testing rates for troponin than White patients (BIDMC: -2.1%, 95% CI: -2.6% to -1.6%, U-M: -2.2%, 95% CI: -2.7% to -1.8%). The observed racial testing differences may impact AI models trained using available laboratory results. The findings also motivate further study of how such differences arise and how to mitigate potential impacts on AI models.

A multi-criteria decision-making model for sustainable selection of coastal protection structures

Although nature-based and hybrid (i.e., a combination of hard and nature-based) solutions provide well-documented benefits to both the environment and people living nearby, they cannot be implemented in all coastal zones. Thus, the selection process of coastal protection structures (CPSs) requires an in-depth and comprehensible assessment with an effective and robust decision-support framework. Existing frameworks in the literature, however, do not focus on the selection of nature-based solutions and environmental and social factors were not considered thoroughly in these frameworks. In this respect, this research aims to develop a decision-support model that guides decision-makers in selecting the most appropriate and sustainable CPSs. The framework was designed in a way that includes a variety of environmental and social factors along with economic and technical aspects. Then, fuzzy AHP and TOPSIS analyses were conducted to determine the relative importance of the selection factors and to select the most appropriate CPS. Factors such as sediment dynamics, geotechnical issues, construction cost, transportation and navigation effects, and technological know-how were among the top-rated factors. TOPSIS analysis results revealed that mangrove forests and artificial reefs were the top two most sustainable CPSs. The proposed framework can minimize the social and environmental impacts of CPSs and expedite the green transition in the coastal environment.

Predicting High-Flow Nasal Cannula Oxygen Therapy Failure in Patients With Acute Hypoxaemic Respiratory Failure Using Machine Learning: Model Development and External Validation.

To develop and validate a prediction model for high-flow nasal cannula (HFNC) failure in patients with acute hypoxaemic respiratory failure (AHRF). AHRF accounts for a major proportion of intensive care unit (ICU) admissions and is associated with high mortality. HFNC is a non-invasive respiratory support technique that can improve patient oxygenation. However, HFNC failure, defined as the need for escalation to invasive mechanical ventilation, can lead to delayed intubation, prolonged mechanical ventilation and increased risk of mortality. Timely and accurate prediction of HFNC failure has important clinical implications. Machine learning (ML) can improve clinical prediction. Multicentre observational study. This study analysed 581 patients from an academic medical centre in Boston and 180 patients from Guangzhou, China treated with HFNC for AHRF. The Boston dataset was randomly divided into a training set (90%, n = 522) and an internal validation set (10%, n = 59), and the model was externally validated using the Guangzhou dataset (n = 180). A random forest (RF)-based feature selection method was used to identify predictive factors. Nine machine learning algorithms were selected to build the predictive model. The area under the receiver operating characteristic curve (AUC) and performance evaluation parameters were used to evaluate the models. The final model included 38 features selected using the RF method, with additional input from clinical specialists. Models based on ensemble learning outperformed other models (internal validation AUC: 0.83; external validation AUC: 0.75). Important predictors of HFNC failure include Glasgow Coma Scale scores and Sequential Organ Failure Assessment scores, albumin levels measured during HFNC treatment, ROX index at ICU admission and sepsis. This study developed an interpretable ML model that accurately predicts the risk of HFNC failure in patients with AHRF. Clinicians and nurses can use ML models for early risk assessment and decision support in AHRF patients receiving HFNC. TRIPOD checklist for prediction model studies was followed in this study. Patients were involved in the sample of the study.

Decision Support Model for Selecting the Optimal Blockchain Oracle Platform: An Evaluation of Key Factors

Smart contract-based applications are executed in a blockchain environment, and they cannot directly access data from external systems, which is required for the service provision of these applications. Instead, smart contracts use agents known as blockchain oracles to collect and provide data feeds to the contracts. The functionality and compatibility with smart contract applications need to be considered when selecting the best-fit oracle platform. As the number of oracle alternatives and their features increases, the decision-making process becomes increasingly complex. Selecting the wrong or sub-optimal oracle is costly and may lead to severe security risks. This paper provides a decision support model for the oracle selection problem. The model supports smart contract decision-makers in selecting a secure, cost-effective, and feasible oracle platform for their applications. We interviewed oracle co-founders and smart contracts experts to refine and validate the decision model. Two real-world smart contract application case studies were used to evaluate the model. Our model prioritises and suggests more than one possible oracle platform based on the developer's required criteria, security assessment, and cost analysis. Moreover, this guided decision model serves to reveal issues that may go unnoticed if done haphazardly, reduce decision-making efforts, and provide a cost-effective solution.

A micro-network within the port for vessel anchorage selection decision support

Upon arrival at ports, selecting the appropriate anchorage is a critical concern shared by ship captains and port managers. This study aims to exploit the rich geographic and semantic information embedded in historical ship trajectory data to construct a micro-network of port elements and employ network analysis techniques to mine spatial service relationships among port elements, thereby establishing an Anchorage Selection Decision Support (ASDS) model. This process involves extracting vessels’ dwell location information within ports from trajectory data and using this information and its sequence to establish an Anchorages-Berths Micro-Network (ABMN) in port. By applying association rule mining techniques, this research reveals the spatial service relationships within the micro-network between anchorages and berths, and integrates relevant indicators to develop the ASDS model. The aim is to help pilots of approaching vessels in making more rational anchorage choices, thus optimizing the operational efficiency of vessels in ports. The effectiveness of this approach has been validated through experiments conducted in the research region at the Nanjing port of the Yangtze River. This research is significant for exploring the correlation between port anchorages and berths, as well as for selecting appropriate anchorages for vessels arriving at the port.

A Decision Support Framework for Resilient and Sustainable Service Design

AbstractResilient and sustainable service design is essential for ensuring the longevity and effectiveness of service systems. However, existing literature often neglects key aspects such as articulating resilience attributes and integrating sustainability dimensions. This study proposes a decision support model for a “resilient-sustainable service design” that merges service design principles with resilient system attributes and organizational sustainability goals. The framework incorporates a multi-objective mathematical programming model and a multi-phased Quality Function Deployment (QFD) approach to derive Pareto optimal solutions using the Brute Force algorithm. Applied in the m-health service sector in Bangladesh, the study reveals significant challenges, including limited awareness of services and logistical inefficiencies. To address these issues, flexible strategies such as demand planning and service innovation are implemented. The findings have direct implications for the improvement of service delivery processes and underscore the importance of considering both resilience and sustainability. While focusing on Bangladesh’s m-health sector, the insights gained have broader relevance globally. The integration of resilience and sustainability principles into service design is crucial for addressing complex challenges across sectors and regions. Future research could involve longitudinal studies to capture evolving resilience strategies and explore resilient-sustainable service systems from a broader perspective. This entails examining various factors such as technological advancements and socio-economic dynamics shaping resilient and sustainable service ecosystems.

Groundwater modelling for decision-support in practice: Insights from Sweden

Groundwater is an essential resource for drinking water, food production, and industrial applications worldwide. Over-exploitation and pollution pose significant risks to groundwater sustainability. Groundwater models can be powerful tools for optimizing use, managing risks, and aiding decision-making. For this purpose, models should assimilate pertinent data and quantify uncertainties in outcomes. We examine applied modelling for characterization and decision support in Sweden from 2010 to 2023. We also review syllabi of water-related courses in Swedish higher education to assess the inclusion and extent of groundwater modelling education. We find that important academic advances in groundwater modelling over the past two decades have not translated into practical application within Sweden’s industry, that uncertainty quantification is rarely undertaken, and that groundwater modelling remains a low priority in higher education. Based on these findings, we offer recommendations that, while informed by the Swedish context, hold relevance for educational institutions, industry, and decision-makers internationally.

Population dynamics of a desert riverine turtle, Pseudemys gorzugi, at the northern edge of its range

Estimating the key demographic parameters of animal populations can enhance our understanding of system dynamics and assist in developing and improving conservation decision–support models. The Rio Grande cooter Pseudemys gorzugi is a conservation reliant freshwater turtle native to lower Rio Grande River Basin (USA and Mexico), with limited knowledge regarding its natural history and population dynamics. In this study, we used seven years of capture–mark–recapture data from the northern edge of the species' range to estimate survival probabilities, changes in abundance, and the probability of transitioning between different size classes while explicitly accounting for the sampling process. We found relatively high survival probabilities across different strata, with large juveniles exhibiting the highest survival (0.98) and small juveniles the lowest (0.71). However, transition probabilities between strata were low, indicating slow somatic growth rates. Our pattern‐oriented modelling revealed a low overall mean estimate of egg survival (0.024), warranting further empirical confirmation. Our study provides the first comprehensive demographic analysis of P. gorzugi encompassing an array of size and sex classes. Overall, we consider the population of P. gorzugi in the Black River robust, highlighting the importance of this river system to the species' persistence in the northern extent of its range, where the population is isolated from its broader distribution. The demographic estimates and ecological insights provided by our study offer critical data for parameterizing decision‐support models to ensure that P. gorzugi conservation strategies are grounded in the best available science.

A Decision Support Model for Lean Supply Chain Management in City Multifloor Manufacturing Clusters

City manufacturing has once again become one of the priority areas for the sustainable development of smart cities thanks to the use of a wide range of green technologies and, first of all, additive technologies. Shortening the supply chain between producers and consumers has significant effects on economic, social, and environmental dimensions. Zoning of city multifloor manufacturing (CMFM) in areas with a compact population in large cities in the form of clusters with their own city logistics nodes (CLNs) creates favorable conditions for promptly meeting the needs of citizens for goods of everyday demand and for passenger and freight transportation. City multifloor manufacturing clusters (CMFMCs) have been already studied quite a lot for their possible uses; nevertheless, an identified research gap is related to supply chain design efficiency concerning urban CMFMCs. Thus, the main objective of this study was to explore the possibilities of lean supply chain management (LSCM) as the integrated application of lean manufacturing (LM) approaches and I4.0 technologies for customer-centric value stream management based on eliminating all types of waste, reducing the use of natural and energy resources, and continuous improvement of processes related to logistics activities. This paper presents a decision support model for LSCM in CMFMCs, which is a mathematical deterministic model. This model justifies the minimization of the number of road transport transfers within the urban area and the amount of stock that is stored in CMFMC buildings and in CLNs, and also regulating supplier lead time. The model was verified and validated using appropriately selected test data based on the case study, which was designed as a typical CMFM manufacturing system with various parameters of CMFMCs and urban freight transport frameworks. The feasibility of using the proposed model for value stream mapping (VSM) and managing logistics processes and inventories in clusters is discussed. The findings can help decisionmakers and researchers improve the planning and management of logistics processes and inventory in clusters, even in the face of unexpected disruptions.

An FMEA decision support model for hotel risk assessment based on the risk tolerance of multi-type travelers and online reviews

As a reliability management tool, failure mode and effect analysis (FMEA) is widely used to solve the problem of risk assessment in all trades and professions. Considering the opinions interaction and the diversity of risk tolerance of experts, this paper constructs a decision-support model of FMEA based on dynamic interaction of opinions and risk tolerance. Firstly, word frequency statistics and Word2Vec analysis are applied on 14,194 online reviews collected from TripAdvisor.com to identify major risk factors. Secondly, the consensus opinion of risk is reached through the dynamic interaction of opinions. Furthermore, the weight problem of risk factors is solved by a linear programming model with the goal of minimum individual regret, and the priority of risk is determined by an improved TODIM method considering risk tolerance. Through simulation experiments, sensitivity analysis and comparative analysis, the effectiveness and superiority of the proposed method are verified, and the theoretical research of FMEA method is enriched. This study provides a direction for the future improvement of hotel services based on understanding the preferences and interactions of multi-type travelers.

Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology.

Rapidly expanding medical literature challenges oncologists seeking targeted cancer therapies. General-purpose large language models (LLMs) lack domain-specific knowledge, limiting their clinical utility. This study introduces the LLM system Medical Evidence Retrieval and Data Integration for Tailored Healthcare (MEREDITH), designed to support treatment recommendations in precision oncology. Built on Google's Gemini Pro LLM, MEREDITH uses retrieval-augmented generation and chain of thought. We evaluated MEREDITH on 10 publicly available fictional oncology cases with iterative feedback from a molecular tumor board (MTB) at a major German cancer center. Initially limited to PubMed-indexed literature (draft system), MEREDITH was enhanced to incorporate clinical studies on drug response within the specific tumor type, trial databases, drug approval status, and oncologic guidelines. The MTB provided a benchmark with manually curated treatment recommendations and assessed the clinical relevance of LLM-generated options (qualitative assessment). We measured semantic cosine similarity between LLM suggestions and clinician responses (quantitative assessment). MEREDITH identified a broader range of treatment options (median 4) compared with MTB experts (median 2). These options included therapies on the basis of preclinical data and combination treatments, expanding the treatment possibilities for consideration by the MTB. This broader approach was achieved by incorporating a curated medical data set that contextualized molecular targetability. Mirroring the approach MTB experts use to evaluate MTB cases improved the LLM's ability to generate relevant suggestions. This is supported by high concordance between LLM suggestions and expert recommendations (94.7% for the enhanced system) and a significant increase in semantic similarity from the draft to the enhanced system (from 0.71 to 0.76, P = .01). Expert feedback and domain-specific data augment LLM performance. Future research should investigate responsible LLM integration into real-world clinical workflows.