Training Scenarios Research Articles

Development and validation of Simulation Scenario Quality Instrument (SSQI)

BackgroundDue to the unmet need for valid instruments that evaluate critical components of simulation scenarios, this research aimed to develop and validate an instrument that measures the quality of healthcare simulation scenarios.MethodsA sequential transformative mixed-method research design was used to conduct the study. The development and validation of the instrument involved two phases: the qualitative phase, which included defining the instrument's theoretical background and instrument construction, followed by the quantitative phase, where the instrument was piloted and validated. The qualitative study included 17 healthcare simulation experts, where three focus group was conducted, and the first version of the instrument was constructed based on the focus group analysis and the theoretical framework constructed using the literature review. During the quantitative phase, the instrument’s quantitative piloting included 125 healthcare simulation scenarios; then, the instrument went through construct validity and reliability testing.ResultsContent experts confirmed the theoretical model and instrument framework. The average item content validity index (I-CVI) scores and the average of the I-CVI scores (S-CVI/Ave) for all items on the scale or the average proportion relevance judged by all experts was 0.87. The conformity factor analysis results showed a good fit for the proposed 10-factor model (CFI (the comparative fit index) = 0.998, Tucker-Lewis index = 0.998, Root mean square error of approximation (RMSEA) = 0.061. The final instrument included ten domains: 1. Learning objectives, 2. Target group, 3. Culture, 4. Scenario case, 5. Scenario narrative briefing, 6. Scenario complexity, 7. Scenario flow, 8. Fidelity, 9. Debriefing, and 10. Assessment. The SSQI included 44 items that are rated on a 3-point scale (Meets Expectations = (2), Needs Improvement, (1), Inadequate (0)).ConclusionThis validated and reliable instrument will be helpful to healthcare educators and simulation experts who want to develop simulation-based training scenarios and ensure the quality of written scenarios.

Personalized virtual reality exposure for panic disorder and agoraphobia: A preliminary neurophysiological study

BackgroundPersonalization is considered an important principle in virtual reality (VR) exposure therapy. We aimed to identify whether personalized VR exposure could provoke increased anxiety in patients with panic disorder and agoraphobia as it is considered the first step in successful treatment for anxiety. MethodsWe performed a double-arm, one-day preliminary study among 28 patients with panic disorder and agoraphobia. Three sessions of VR exposure, including a theater, train, and elevator scenario, were conducted in two groups. In the personalized group (n = 14), the brightness and crowd density were customized based on a pre-assessment. In the control group (n = 14), these conditions were fully randomized. Self-reported anxiety, heart rate, skin conductance, and electroencephalography were measured before, during, and after the VR sessions. ResultsIn the later VR sessions, higher self-reported anxiety levels measured by the Visual Analogue Scale were observed in the personalized exposure group. Increased heart rates during and after the VR sessions were observed in the personalized group. The changes in skin conductance peaks were not significantly different between the groups, but the increase in skin conductance was associated with the participants' perception of presence. The electroencephalogram showed widespread increases in alpha waves in the frontal and temporal areas of the brain in the personalized group than in the control group. ConclusionPersonalized VR exposure elicits stronger anxiogenic effects in patients with panic disorder and agoraphobia as suggested by self-report and neurophysiological data. Personalization of VR exposure has the potential for effective behavioral therapy.

Worst Case Scenarios in Soccer Training and Competition: Analysis of Playing Position, Congested Periods, and Substitutes

ABSTRACT Aim: To understand mean (WCSmean) and peak (WCSpeak) worst case scenarios within training and game play in male professional soccer. Methods: Thirty-one (n = 31) first team players were monitored across 37 matches and 14 MD-3 sessions. Playing status was distinguished, football drills analyzed, and performance explored in long-period: >6 days, moderate-period: 5–6 days, and congested-period: ≤4 days. Relative total distance (TD), high-speed running distance (HSRD, >19.8 km·h−1), sprint distance (SD, >25.2 km·h−1), accelerations/decelerations (A+D, >3 m·s−2), accelerations (Acc, >3 m·s−2), and decelerations (Dec, >−3 m·s−2) were measured as well as Maximum acceleration (Max Acc; m·s−2) and deceleration (Max Dec; m·s−2). Results: Analysis of variance found differences between matches and training in WCSmean for TD, HSRD, SD, and Max Dec in all positions (p < .001; partial η2 > .275). Fullbacks displayed differences between match and training in Max Acc (moderate ESs; p < .001), while center backs and central midfielders in Max Dec (large ESs; p > .05). Main effects of playing status were discovered for all metrics except Max Dec (p < .001; partial η2 > .124). Analysis showed differences between long- and congested-period for A+D and Dec (large ESs; p ≤ .05). Conclusions: Findings provide more insights into short peak intensity demands of soccer showing that the maximum high velocity action of acceleration and deceleration is not being replicated in training. Nonstarters lack maximum intensity exposure in matches (WCSpeak) increasing the gap between training and competition even higher during congested fixture periods.

The influence of human-computer interface on usability and technology acceptance of VR-based shooting training with a comparison with typical shooting range

This paper investigates the influence of the human-computer interface on the usability and technology acceptance of VR-based shooting training in comparison to a typical shooting range. We conducted an experiment with 160 active duty State Protection Service officers who performed shooting tasks in real and four different VR environments with different levels of complexity. The results showed that the interface and the type of training scenario had a significant effect on the usability and technology acceptance of VR-based shooting training. Participants also preferred the VR environment to the real environment for shooting training, because it provided more immersion and motivation. Surprisingly, as the complexity of the virtual environment increased, usability and intention to use decreased. The paper discusses the implications of these findings for the design and development of VR-based shooting training systems.

Establishing Phase Definitions for Jump and Drop Landings and an Exploratory Assessment of Performance-Related Metrics to Monitor During Testing.

Harry, JR, Simms, A, and Hite, M. Establishing phase definitions for jump and drop landings and an exploratory assessment of performance-related metrics to monitor during testing. J Strength Cond Res 38(2): e62-e71, 2024-Landing is a common task performed in research, physical training, and competitive sporting scenarios. However, few have attempted to explore landing mechanics beyond its hypothesized link to injury potential, which ignores the key performance qualities that contribute to performance, or how quickly a landing can be completed. This is because a lack of (a) established landing phases from which important performance and injury risk metrics can be extracted and (b) metrics known to have a correlation with performance. As such, this article had 2 purposes. The first purpose was to use force platform data to identify easily extractable and understandable landing phases that contain metrics linked to both task performance and overuse injury potential. The second purpose was to explore performance-related metrics to monitor during testing. Both purposes were pursued using force platform data for the landing portion of 270 jump-landing trials performed by a sample of 14 NCAA Division 1 men's basketball players (1.98 ± 0.07 m; 94.73 ± 8.01 kg). The proposed phases can separate both jump-landing and drop-landing tasks into loading, attenuation, and control phases that consider the way vertical ground reaction force (GRF) is purposefully manipulated by the athlete, which current phase definitions fail to consider. For the second purpose, Pearson's correlation coefficients, the corresponding statistical probabilities ( α = 0.05), and a standardized strength interpretation scale for correlation coefficients (0 < trivial ≤ 0.1 < small ≤ 0.3 < moderate ≤ 0.5 < large ≤ 0.7 < very large) were used for both the group average (i.e., all individual averages pooled together) and individual data (i.e., each individual's trials pooled together). Results revealed that landing time, attenuation phase time, average vertical GRF during landing, average vertical GRF during the attenuation phase, average vertical GRF during the control phase, vertical GRF attenuation rate, and the amortization GRF (i.e., GRF at zero velocity) significantly correlated with landing performance, defined as the ratio of landing height and landing time ( R ≥ ± 0.58; p < 0.05), such that favorable changes in those metrics were associated with better performance. This work provides practitioners with 2 abilities. First, practitioners currently assess jump capacity using jump-landing tests (e.g., countermovement jump) with an analysis strategy that makes use of landing data. Second, this work provides preliminary data to guide others when initially exploring landing test results before identifying metrics chosen for their own analysis.

Authoring Moving Parts of Objects in AR, VR and the Desktop

Creating digital twins of real objects is becoming more popular, with smartphones providing 3D scanning capabilities. Adding semantics to the reconstructed virtual objects is important to possibly reproduce training scenarios that otherwise could demand significant resources or happen in dangerous scenarios in some cases. The aim of this work is to evaluate the usability of authoring object component behaviors in immersive and non-immersive approaches. Therefore, we present an evaluation of the perceived ease of use to author moving parts of objects under three different conditions: desktop, augmented reality (AR) and virtual reality (VR). This paper provides insights into the perceived benefits and issues that domain experts might encounter when authoring geometrical component behaviors across each interface. A within-subject study is the major contribution of this work, from which is presented an analysis based on the usability, workload and user interface preferences of participants in the study. To reduce confounding variables in the study, we ensured that the virtual objects and the environment used for the evaluation were digital twins of the real objects and the environment that the experiment took place in. Results show that the desktop interface was perceived as more efficient and easier to use based on usability and workload measures. The desktop was preferred for performing component selection but no difference was found in the preference for defining a behavior and visualizing it. Based on these results, a set of recommendations and future directions are provided to achieve a more usable, immersive authoring experience.

Plausibility of BBR as CUBIC’S Replacement and proposed improvement to BBR using GENET

Recent advancements in deep reinforcement learning (RL) have opened new avenues for enhancing network congestion control algorithms. Our research builds upon these developments, particularly focusing on the BBR (Bottleneck Bandwidth and Round-trip propagation time) congestion control algorithm. We propose integrating GENET's reinforcement learning framework, a novel training paradigm that has demonstrated success in various network adaptation algorithms, including adaptive video streaming, congestion control, and load balancing. GENET leverages curriculum learning to effectively train RL models by progressively introducing more challenging network environments. This method counters the common pitfalls in RL training, such as suboptimal performance in a wide range of environments and poor generalization in narrowly defined training scenarios. Our approach exploits the strengths of GENET in identifying and emphasizing network conditions where the current RL model underperforms compared to traditional rule-based baselines, thereby facilitating significant improvements. This research aims to demonstrate that applying GENET's methodology to the BBR congestion control algorithm can yield RL policies that surpass both regularly trained RL policies and conventional baselines, thereby advancing the efficiency and reliability of network congestion control.

Research on the Innovation of English Teaching Mode Based on Artificial Intelligence

This article presents a research on innovative approaches in English language teaching through the application of Artificial Intelligence (AI). The aim is to address the issues found in traditional English language teaching methods by utilizing AI technology. The article begins by identifying the problems in traditional English language teaching and exploring the current applications of AI in education. Then, it specifically discusses the application of AI technology in English language teaching, such as interactive listening training, intelligent error correction and personalized post-class guidance, and cloud-based intelligent English translation teaching. Subsequently, a proposal is put forward to construct a new English language teaching system based on AI technology, including modules for engaging explanations, interactive English assignments, and creating scenarios for cross-cultural communication oral training. Lastly, an evaluation method for assessing the innovation of English language teaching models based on AI is introduced, which includes feedback and perceptions from both teachers and students, as well as an assessment of teaching outcomes. The research findings of this article can provide a more intelligent, personalized, and effective English language teaching approach, thus positively impacting the quality of English language teaching and enhancing students’ interest in learning.

A holistic solution to virtual coupling based urban rail train control system

Virtual coupling (VC) is an emerging concept that is gaining popularity in railway control systems. It has the potential to significantly enhance railway capacity and flexibility, and has attracted considerable research attention. This study explores several key aspects of VC research in the context of urban rail transit systems, including system structure, train platoon control method, scenario definition, and train scheduling. The paper presents three main contributions: Firstly, it proposes a potential system structure for VC and discusses infrastructure occupancy procedures within typical operational scenarios. Secondly, a VC platoon controller based on the adaptive back-stepping method is designed to realize the function of the operation layer. Thirdly, combining operational scenario definition and control methods, we propose a VC train scheduling method based on mixed integer linear programming (MILP). A case study demonstrated that the proposed solution could effectively achieve the primary functions of VC.

Deep learning assisted three triboelectric driving operation sensors for driver training and behavior monitoring

Improving the driving skills of drivers, particularly during the training stage, is crucial in reducing the likelihood of road traffic accidents. In this work, a driver training assistance system (DTAS) is developed for driver training and behavior monitoring. The DTAS integrates three triboelectric driving operation sensors, including gear shift sensor, steering angle sensor, and pedal sensors. Through the ingenious structural design of contact-separation and freestanding-triboelectric-layer mode, these triboelectric sensors have the characteristics of simple structure, easy manufacture and installation, and self-powered, which avoids the complex wiring problem in the limited space of the vehicle. The basic electrical performance test of triboelectric sensors and driving simulation experiment show that the developed DTAS can monitor the driver behavior and provide the feedback on each driver’s operation process in real-time. Combined with deep learning (DL) technology, the DTAS can identify whether the driving operation of drivers in specific training scenarios is correct or not, with an accuracy rate of 97.5%. This work is aimed at assisting the novice or learner drivers in driving training, which helps to improve their driving skills and form good driving habits. The proposed scheme can provide new ideas for the innovative exploration of driving training modes without coaching.

Improving mixed-integer temporal modeling by generating synthetic data using conditional generative adversarial networks: A case study of fluid overload prediction in the intensive care unit

ObjectiveThe challenge of mixed-integer temporal data, which is particularly prominent for medication use in the critically ill, limits the performance of predictive models. The purpose of this evaluation was to pilot test integrating synthetic data within an existing dataset of complex medication data to improve machine learning model prediction of fluid overload. Materials and methodsThis retrospective cohort study evaluated patients admitted to an ICU ≥ 72 h. Four machine learning algorithms to predict fluid overload after 48–72 h of ICU admission were developed using the original dataset. Then, two distinct synthetic data generation methodologies (synthetic minority over-sampling technique (SMOTE) and conditional tabular generative adversarial network (CTGAN)) were used to create synthetic data. Finally, a stacking ensemble technique designed to train a meta-learner was established. Models underwent training in three scenarios of varying qualities and quantities of datasets. ResultsTraining machine learning algorithms on the combined synthetic and original dataset overall increased the performance of the predictive models compared to training on the original dataset. The highest performing model was the meta-model trained on the combined dataset with 0.83 AUROC while it managed to significantly enhance the sensitivity across different training scenarios. DiscussionThe integration of synthetically generated data is the first time such methods have been applied to ICU medication data and offers a promising solution to enhance the performance of machine learning models for fluid overload, which may be translated to other ICU outcomes. A meta-learner was able to make a trade-off between different performance metrics and improve the ability to identify the minority class.

A machine learning diagnosis of the severe accident progression

A Machine Learning (ML) platform is proposed to assist the operator in diagnosing the severe accident progression, where some of the signal is corrupted, and/or estimation of key parameters for accident management is needed. To predict the key parameters for accident management such as lost signals for the pressure and core water level, the amount of hydrogen generation, and the mass of nuclear aerosols, a Machine Learning (ML) platform with long short term memory (LSTM) network is proposed, where multiple accident scenarios can be accumulated and used for training. Training and test data were produced by MELCOR simulations of the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident at unit 3. The MELCOR simulation data is presented as time series at equal intervals. Feature variables were selected among many plant parameters, where the importance ranking was determined by a recursive feature elimination technique using RandomForestRegressor. We performed a sensitivity study of the different choices of training and test scenarios, the number of feature variables, the number of neurons, and target variables on the prediction accuracy of the ML platform. Among eight MELCOR simulation results, different combinations of training and test data are selected in terms of similarity between them. It is found that when the number of feature variables is more than five, the proposed ML platform was able to predict not only similar test data but also unseen test data with reasonable accuracy. It is also found that the proposed ML platform consistently predicts any of the lost signals and key parameters for accident management with reasonable accuracy. Pre-trained ML model predicted the target variable within several minutes on a desktop computer.

Deep-Reinforcement-Learning-Based Low-Carbon Economic Dispatch for Community-Integrated Energy System under Multiple Uncertainties

A community-integrated energy system under a multiple-uncertainty low-carbon economic dispatch model based on the deep reinforcement learning method is developed to promote electricity low carbonization and complementary utilization of community-integrated energy. A demand response model based on users’ willingness is proposed for the uncertainty of users’ demand response behavior; a training scenario set of a reinforcement learning agent is generated with a Latin hypercube sampling method for the uncertainties of power, load, temperature, and electric vehicle trips. Based on the proposed demand response model, low-carbon economic dispatch of the community-integrated energy system under multiple uncertainties is achieved by training the agent to interact with the environment in the training scenario set and reach convergence after 250 training rounds. The simulation results show that the reinforcement learning agent achieves low-carbon economic dispatch under 5%, 10%, and 15% renewable energy/load fluctuation scenarios, temperature fluctuation scenarios, and uncertain scenarios of the number of trips, time periods, and mileage of electric vehicles, with good generalization performance under uncertain scenarios.

Virtual Scenarios of Earthquake Early Warning to Disaster Management in Smart Cities Based on Auxiliary Classifier Generative Adversarial Networks.

Effective response strategies to earthquake disasters are crucial for disaster management in smart cities. However, in regions where earthquakes do not occur frequently, model construction may be difficult due to a lack of training data. To address this issue, there is a need for technology that can generate earthquake scenarios for response training at any location. We proposed a model for generating earthquake scenarios using an auxiliary classifier Generative Adversarial Network (AC-GAN)-based data synthesis. The proposed ACGAN model generates various earthquake scenarios by incorporating an auxiliary classifier learning process into the discriminator of GAN. Our results at borehole sensors showed that the seismic data generated by the proposed model had similar characteristics to actual data. To further validate our results, we compared the generated IM (such as PGA, PGV, and SA) with Ground Motion Prediction Equations (GMPE). Furthermore, we evaluated the potential of using the generated scenarios for earthquake early warning training. The proposed model and algorithm have significant potential in advancing seismic analysis and detection management systems, and also contribute to disaster management.

Token incentive framework for virtual-reality-based construction safety training

Virtual reality (VR) and blockchain are two emerging technologies with significant potential for the construction industry. Numerous studies have investigated the use of VR to improve construction safety training. However, there is a lack of research focused on encouraging employees to participate in VR training. Fortunately, incentive mechanisms are a powerful motivator, capable of driving employees to actively participate in and excel at safety practices. Incorporating blockchain-based incentives into safety training creates a win-win scenario which can substantially improve training outcomes. To this end, this paper proposes a token incentive mechanism based on blockchain technology for VR-based safety training, consisting of unique safety training scenarios and a flexible virtual environment that enables users to acquire knowledge of construction safety regulations. The framework supports real-time interaction and communication among users and provides blockchain-based tokens to incentivize safe conduct and reward individuals who have undergone safety training. A prototype was developed and evaluated through interactive system trials with senior students, site managers, and construction site employees. Results demonstrate that the proposed framework can motivate users to participate in safety training regularly and improve their safety awareness and knowledge before performing tasks at construction sites. In addition, the framework fosters collaboration among employees and assists safety managers in creating novel VR-based safety training scenarios that can be delivered to workers easily. By incentivizing safety training, the proposed framework contributes to establishing a new safety culture wherein workers can effectively acquire knowledge and financial benefits through completing safety training activities.

The combination of listening to background music and deep breathing as a stress reduction strategy during dental preclinical study.

According to our earlier study, background music can help foster dental preclinical manual dexterity training and learning and has the potential to increase student satisfaction and productivity. Deep diaphragmatic breathing helps produce a relaxed mental state and is considered a behavioral coping strategy to reduce anxiety. Experimental data on whether background music combined with deep belly breathing reduces the stress and anxiety of preclinical dental students during practice is lacking. Thus, we focused on evaluating the effects of the combined use of these two techniques on dental students during preclinical manual skills training. Thirty-one voluntary 3rd-year dental students' anxiety levels and their behavioral changes were monitored via anonymous, voluntary surveys including the Patient Health Questionnaire-4, General Health Questionnaire-12 (GHQ-12), and Visual Analogue Scale for Anxiety (VAS-A). Overall, stress measured by GHQ-12 decreased significantly in female students (p=0.025), but not in males. Both background music and deep belly breathing significantly reduced all students' fear and stress based on their VAS-A scores in practice. However, a clear difference from the nonuse of coping strategies could not be conclusively shown. The students regarded the course positively based on survey replies. The combined use of the two coping strategies delivered beneficial effects to the preclinical dental students: they reduced the amount of stress in the supervised preclinical students and considerably cut down the students' anxiety. Further investigation of this new dental training scenario may help professionals provide better student education and care management during the dental preclinical phase.

3D surgical instrument collection for computer vision and extended reality

The availability of computational hardware and developments in (medical) machine learning (MML) increases medical mixed realities’ (MMR) clinical usability. Medical instruments have played a vital role in surgery for ages. To further accelerate the implementation of MML and MMR, three-dimensional (3D) datasets of instruments should be publicly available. The proposed data collection consists of 103, 3D-scanned medical instruments from the clinical routine, scanned with structured light scanners. The collection consists, for example, of instruments, like retractors, forceps, and clamps. The collection can be augmented by generating likewise models using 3D software, resulting in an inflated dataset for analysis. The collection can be used for general instrument detection and tracking in operating room settings, or a freeform marker-less instrument registration for tool tracking in augmented reality. Furthermore, for medical simulation or training scenarios in virtual reality and medical diminishing reality in mixed reality. We hope to ease research in the field of MMR and MML, but also to motivate the release of a wider variety of needed surgical instrument datasets.

High-Quality and Reproducible Automatic Drum Transcription from Crowdsourced Data

Within the broad problem known as automatic music transcription, we considered the specific task of automatic drum transcription (ADT). This is a complex task that has recently shown significant advances thanks to deep learning (DL) techniques. Most notably, massive amounts of labeled data obtained from crowds of annotators have made it possible to implement large-scale supervised learning architectures for ADT. In this study, we explored the untapped potential of these new datasets by addressing three key points: First, we reviewed recent trends in DL architectures and focused on two techniques, self-attention mechanisms and tatum-synchronous convolutions. Then, to mitigate the noise and bias that are inherent in crowdsourced data, we extended the training data with additional annotations. Finally, to quantify the potential of the data, we compared many training scenarios by combining up to six different datasets, including zero-shot evaluations. Our findings revealed that crowdsourced datasets outperform previously utilized datasets, and regardless of the DL architecture employed, they are sufficient in size and quality to train accurate models. By fully exploiting this data source, our models produced high-quality drum transcriptions, achieving state-of-the-art results. Thanks to this accuracy, our work can be more successfully used by musicians (e.g., to learn new musical pieces by reading, or to convert their performances to MIDI) and researchers in music information retrieval (e.g., to retrieve information from the notes instead of audio, such as the rhythm or structure of a piece).

Synthetic data generation for the continuous development and testing of autonomous construction machinery

Abstract The development and testing of autonomous systems require sufficient meaningful data. However, generating suitable scenario data is a challenging task. In particular, it raises the question of how to narrow down what kind of data should be considered meaningful. Autonomous systems are characterized by their ability to cope with uncertain situations, i.e. complex and unknown environmental conditions. Due to this openness, the definition of training and test scenarios cannot be easily specified. Not all relevant influences can be sufficiently specified with requirements in advance, especially for unknown scenarios and corner cases, and therefore the “right” data, balancing quality and efficiency, is hard to generate. This article discusses the challenges of automated generation of 3D scenario data. We present a training and testing loop that provides a way to generate synthetic camera and Lidar data using 3D simulated environments. Those can be automatically varied and modified to support a closed-loop system for deriving and generating datasets that can be used for continuous development and testing of autonomous systems.

TCCC Decision Support With Machine Learning Prediction ofHemorrhage Risk, Shock Probability.

Expected future delays in evacuation during near-peer conflicts in remote locales are expected to require extended care including prolonged field care over hours to days. Such delays can increase potential complications, such as insufficient blood flow (shock), bloodstream infection (sepsis), internal bleeding (hemorrhage), and require more complex treatment beyond stabilization. The Trauma Triage Treatment and Training Decision Support (4TDS) system is a real-time decision support system to monitor casualty health and identify such complications. The 4TDS software prototype operates on an Android smart phone or tablet configured for use in the DoD Nett Warrior program. It includes machine learning models to evaluate trends in six vital signs streamed from a sensor placed on a casualty to identify shock probability, internal hemorrhage risk, and need for a massive transfusion. The project team used a mixed methods approach to create and evaluate the system including literature review, rapid prototyping, design requirements review, agile development, an algorithm "silent test," and usability assessments with novice to expert medics from all three services. Both models, shock (showing an accuracy of 0.83) and hemorrhage/massive transfusion protocol, were successfully validated using externally collected data. All usability assessment participants completed refresher training scenarios and were able to accurately assess a simulated casualty's condition using the phone prototype. Mean responses to statements on evaluation criteria [e.g., fit with Tactical Combat Casualty Care (TCCC), ease of use, and decision confidence] fell at five or above on a 7-point scale, indicating strong support. Participatory design ensured 4TDS and machine learning models reflect medic and clinician mental models and work processes and built support among potential users should the system transition to operational use. Validation results can support 4TDS readiness for FDA 510k clearance as a Class II medical device.