Quantitative Performance Measures Research Articles

Hop to It! A Systematic Review and Longitudinal Meta-analysis of Hop Performance After ACL Reconstruction.

Hop testing is widely used by clinicians to monitor rehabilitation and decide when to return to sport following anterior cruciate ligament reconstruction (ACLR); however, the trajectory of long-term hop performance has not been summarised. To investigate hop performance change over time after ACLR. Systematic review with longitudinal meta-analysis. MEDLINE, EMBASE, CINAHL, Scopus, Cochrane CENTRAL and SPORTDiscus to 28 February 2023. Studies with ≥ 50 participants following primary ACLR, with mean participant age of 18-40years, reporting a quantitative measure of hop performance (e.g. single forward hop distance). Results had to be reported for the ACLR limb and compared with (1) the contralateral limb (within person) and/or (2) an uninjured control limb (between person). We included 136 studies of 23,360 participants. Performance was similar across different hop tests, with steep initial improvements in within-person symmetry, tailing off after 18-24months. ACLR limb hop performance was 5-10% lower compared with the contralateral limb at 1year post-surgery, with largest deficits observed for vertical hop [87.0% contralateral limb (95% CI 85.3-88.8) compared with single forward hop 93.8% (95% CI 92.8-94.9)]. By 3-5years, results were similar between ACLR and contralateral limbs. There were limited data for between-person comparisons (n = 17 studies). Exploratory analyses showed deficits in all forward hopping tests to be very strongly correlated with each other [e.g. single forward and triple hop rho = 0.96 (95% CI 0.90-0.99)], though there was discordance in the relationship between single forward hop and vertical hop performance [rho = 0.27 (95% CI - 0.53 to 0.79)]. Hop performance is comparable to the uninjured limb by 3-5years post-ACLR, with the greatest deficits in within-person symmetry present in vertical and side hop tests. Assessment of hopping in multiple planes and comparison with uninjured controls, may provide the most complete evaluation of functional performance.

A new method to evaluate fluoroscopic system collimator performance.

Fluoroscopy uses collimators to limit the radiation field size. Collimators are often evaluated annually during equipment performance evaluations to maintain compliance with regulatory and/or accreditation bodies. A method to evaluate and quantify fluoroscopy collimator performance was developed. A radiation field and displayed image measurement device consisting of radiopaque rulers and radiochromic film strips was placed on the x-ray source assembly exit window to evaluate fluoroscopy collimator performance. This method was used to evaluate collimator performance on 79 fluoroscopic imaging systems including fixed C-arms, mobile C-arms, mini C-arms, and radiographic fluoroscopic systems. The excess length (EL), excess width (EW), and sum EL + EW of the radiation field relative to the displayed image were measured and compared to the limits specified in 21CFR1020.32. Four systems exceeded these limits. Placing the radiation measurement device at the x-ray source assembly exit window relative to the image receptor cover increased the film exposure rate by a factor up to 14.6. The time required to set up and complete the fluoroscopy collimator performance measurements using this method ranged from 5 to 10 min. This method provides an easily implemented quantitative measure of fluoroscopy system collimator performance that satisfies regulatory and accreditation body requirements.

Secure Efficiency: Navigating Performance Challenges in Multi-Tenant Cloud Security Implementations

Multi-tenant cloud environments offer significant advantages in terms of cost-efficiency and scalability, but they also present unique challenges in balancing robust security measures with optimal performance. This article examines the intricate relationship between security implementations and system performance in shared cloud infrastructures. Through a comprehensive analysis of data isolation techniques, access control mechanisms, network security protocols, and threat detection systems, we identify key areas where security measures can impact performance metrics such as latency, throughput, and resource utilization. Our article employs a mixed-methods approach, combining quantitative performance measurements with qualitative case studies from industry-leading cloud service providers. The findings reveal that while stringent security measures often introduce performance overhead, strategic implementation and optimization can significantly mitigate these effects. We propose a framework for dynamically balancing security and performance requirements, incorporating emerging technologies such as AI-driven threat detection and automated resource allocation. This article contributes to the growing body of knowledge on cloud computing optimization and provides practical insights for cloud architects and security professionals seeking to enhance both the security posture and performance efficiency of multi-tenant cloud environments.

Insights into enhanced learning through virtual reality

IntroductionThis study investigates the impact of Immersive Virtual Reality (VR) on learning outcomes in radiography training using Virtual Medical Coaching's X-Ray Pro VR software. VR offers immersive experiential learning that may enhance academic performance and professional preparedness. MethodsThe study involved 130 students from two consecutive cohorts: one using traditional training methods and the other using VR software. We employed a mixed-methods approach, integrating quantitative measures of academic performance and clinical competency with qualitative insights into student engagement, enjoyment, and stress levels. The VR group used Virtual Medical Coaching simulation software, while the traditional group used physical simulation machines (Siemens Healthineers) to practice positioning anthropomorphic manikins, adjusting exposure settings, and handling X-ray equipment. Data were collected at four points: immediately post-training and at 1 month, 3 months, and 8 months follow-up. ResultsThe VR cohort showed significantly higher knowledge retention, work readiness, and lower stress levels compared to the traditional cohort. Among the participants trained in VR, a vast majority (59 out of 65) reported a high level of satisfaction with the immersive nature of the training, emphasizing its realism and direct applicability to clinical scenarios. Quantitative analysis revealed superior performance metrics for the VR group across all measured time points (p < 0.001). Qualitative feedback highlighted enhanced engagement and enjoyment among VR-trained students. ConclusionThe results of this study clearly indicate that VR training significantly enhances learning outcomes in radiography education. The VR cohort exhibited higher knowledge retention, work readiness, student engagement, and enjoyment, along with lower stress levels and higher confidence compared to the traditional cohort. These findings support the integration of VR into professional training curricula to improve student performance and engagement.

Organizational Commitment as an Intervening Variable in the Effect of Empowerment, Incentives and Self Efficacy on Employee Performance

Badan Kepegawaian Negara (BKN) are an organization of government agencies operating in the field of public service with a unique context as well as regulations and policies on public service management. The high performance of officials is crucial to government agencies. Staff performance is a qualitative and quantitative measure of work performance and work outcomes based on the standard of work achieved by employees per unit of time in performing work in accordance with the assigned responsibilities. The study aims to look at the impact of empowerment, incentives, self-effectiveness, and organizational commitments on staff performance, the influence of empowering, incentive, and self effectiveness in relation to organization commitments, as well as the effect of empoverment, motivation, and Self-effictiveness on employee performance with the intervening variable of organizational engagement. This is a quantitative descriptive study, with respondents of 90 BKN employees, using questionnaires as a basis for data assumption as well as data analysis using SmartPLS software. The results of the analysis show that empowerment, incentives, self efficacy, and commitment of the organization have a positive and significant impact on the performance of the staff, the influence of empowering, incentive, and self effecacy have a significant and positive effect on the organization's commitment, and empowerement, the incentive has an impact on staff's performance with the commitment organiassi as a variable part mediation, and the self efficiency on staff performance with organizational commitmentasi as a full mediation variable.

Pangenomic landscapes shape performances of a synthetic genetic circuit across Stutzerimonas species.

Engineering identical genetic circuits into different species typically results in large differences in performance due to the unique cellular environmental context of each host, a phenomenon known as the "chassis-effect" or "context-dependency". A better understanding of how genomic and physiological contexts underpin the chassis-effect will improve biodesign strategies across diverse microorganisms. Here, we combined a pangenomic-based gene expression analysis with quantitative measurements of performance from an engineered genetic inverter device to uncover how genome structure and function relate to the observed chassis-effect across six closely related Stutzerimonas hosts. Our results reveal that genome architecture underpins divergent responses between our chosen non-model bacterial hosts to the engineered device. Specifically, differential expression of the core genome, gene clusters shared between all hosts, was found to be the main source of significant concordance to the observed differential genetic device performance, whereas specialty genes from respective accessory genomes were not significant. A data-driven investigation revealed that genes involved in denitrification and components of trans-membrane transporter proteins were among the most differentially expressed gene clusters between hosts in response to the genetic device. Our results show that the chassis-effect can be traced along differences among the most conserved genome-encoded functions and that these differences create a unique biodesign space among closely related species.IMPORTANCEContemporary synthetic biology endeavors often default to a handful of model organisms to host their engineered systems. Model organisms such as Escherichia coli serve as attractive hosts due to their tractability but do not necessarily provide the ideal environment to optimize performance. As more novel microbes are domesticated for use as biotechnology platforms, synthetic biologists are urged to explore the chassis-design space to optimize their systems and deliver on the promises of synthetic biology. The consequences of the chassis-effect will therefore only become more relevant as the field of biodesign grows. In our work, we demonstrate that the performance of a genetic device is highly dependent on the host environment it operates within, promoting the notion that the chassis can be considered a design variable to tune circuit function. Importantly, our results unveil that the chassis-effect can be traced along similarities in genome architecture, specifically the shared core genome. Our study advocates for the exploration of the chassis-design space and is a step forward to empowering synthetic biologists with knowledge for more efficient exploration of the chassis-design space to enable the next generation of broad-host-range synthetic biology.

Quantitative measurement of thermal performance of a cylindrical lithium-ion battery

Measuring the thermal performance of lithium-ion battery cells is a critical task in the thermal design of electric vehicle battery packs. This study introduces a quantitative method to assess the thermal performance of cylindrical 21,700 cells considering heat loss, under conditions of both high and low temperature-rises. By supervising the cell’s outgoing heat-flux (heat loss), we thoroughly analyzed the differences in cell heat generation rates between high temperature-rise (HTR) and low temperature-rise (LTR) conditions. The results show that under LTR conditions, the cell heat generation exhibits a “fast-slow-fast” increasing trend, while under HTR conditions, it displays a U-shaped pattern. Notably, the mean cell heat generation rate increases with decreasing temperatures and increasing discharge rates, especially under LTR conditions, where it significantly outperforms that under HTR conditions. This method provides valuable insights for optimizing the thermal design of battery packs.

Exploring entry-exit correlation coefficient (EEC) as new quantitative social activities performance measurement of public space: The case of Dilworth Park, Philadelphia

This study explores a new public space performance measure by integrating public space and walking research to examine the relationship between pedestrian traffic volume and social activity intensity in parks and plazas. Introducing the entry-exit correlation coefficient (EEC), the study utilized bootstrap resampling methods to analyze data and compare to 59 manual observations. Findings reveal a statistically significant negative relationship between EEC and social activities, with a coefficient of −0.2. Lower EEC values, indicating closer correlation between entry and exit, were associated with increased social activities in Dilworth Park. The high synchronization observed between EEC and social activities underscores the potential of EEC as a complementary tool to manual observation in public space research. Additionally, the study emphasizes the importance of considering thermal comfort in monitoring public spaces using EEC. While the study focused on a single case study, its results lay the groundwork for future research to apply automated pedestrian counting and EEC as performance measurements for public spaces.

Meta‐analysis and systematic review for anomaly network intrusion detection systems: Detection methods, dataset, validation methodology, and challenges

AbstractIntrusion detection systems built on artificial intelligence (AI) are presented as latent mechanisms for actively detecting fresh attacks over a complex network. The authors used a qualitative method for analysing and evaluating the performance of network intrusion detection system (NIDS) in a systematic way. However, their approach has limitations as it only identifies gaps by analysing and summarising data comparisons without considering quantitative measurements of NIDS's performance. The authors provide a detailed discussion of various deep learning (DL) methods and explain data intrusion networks based on an infrastructure of networks and attack types. The authors’ main contribution is a systematic review that utilises meta‐analysis to provide an in‐depth analysis of DL and traditional machine learning (ML) in notable recent works. The authors assess validation methodologies and clarify recent trends related to dataset intrusion, detected attacks, and classification tasks to improve traditional ML and DL in NIDS‐based publications. Finally, challenges and future developments are discussed to pose new risks and complexities for network security.

Equilibrium optimiser tuned frequency-shifted internal model control proportional-derivative decoupled dual-loop design for industrial plants followed by experimental validation

Integrating and unstable processes require rugged control demand by virtue of their non-self-regulation character. The occurrence of system poles located at the origin and on the right half of the complex frequency plane causes instability in an open-loop configuration. When these processes proceed with dead time, it demands more sophisticated control requirements. Therefore, binal-loop control schemes are recommended over single-loop controller schemes. For plants with prevalent unstable and integrating dynamics and dead time, a novel modified inferred internal model control-proportional derivative decoupled binal-loop control system is suggested. The inner-loop is controlled by the stabilising proportional derivative controller with Routh–Hurwitz stability constraints. The outer-loop contains an inferred internal model controller for reference-point tracking. The equilibrium optimiser is then used to minimise the integral squared error by tuning the inner and outer-loop controller settings in the confined search space. The suggested strategy delivers appreciable enhancement in the quantitative performance measures as compared to some of the contemporary control techniques. The effect of time-varying disturbance and robust stability analysis is also presented. Finally, utilising a magnetic levitation laboratory setup, the suggested method is experimentally verified using the hardware-in-loop method.

Testing the shock protection performance of Type I construction helmets using impactors of different masses.

Wearing protective helmets is an important prevention strategy to reduce work-related traumatic brain injuries. The existing standardized testing systems are used for quality control and do not provide a quantitative measure of the helmet performance. To analyze the failure characterizations of Type I industrial helmets and develop a generalized approach to quantify the shock absorption performance of Type I industrial helmets based on the existing standardized setups. A representative basic Type I construction helmet model was selected for the study. Top impact tests were performed on the helmets at different drop heights using two different impactor masses (3.6 and 5.0 kg). When the helmets were impacted with potential impact energies smaller than the critical potential impact energy values, there was a consistent relationship between the peak impact force and the potential impact energy. When the helmets were impacted under potential impact energies greater than the critical potential impact energy values, the peak impact forces increased steeply with increasing potential impact energy. A concept of safety margin for construction helmets based on potential impact energy was introduced to quantify the helmets' shock absorption performance. The proposed method will help helmet manufacturers improve their product quality.

GAN based approaches for self-supervised segmentation: A comparative study

Image segmentation is a fundamental image processing technique that involves di-viding an image into distinct regions or segments to enable the analysis and extrac-tion of valuable information. It finds applications in various fields including medi-cine, pattern recognition, computer vision, and security surveillance. Different types of image segmentation techniques can be employed based on specific application requirements, including thresholding, region-based segmentation, edge-based segmen-tation, clustering-based segmentation, active contour models (Snakes), graph-based segmentation, watershed segmentation, and deep learning-based segmentation. the latter has become a powerful tool in image segmentation. Deep learning (DL) models can be trained on large datasets of images, allowing them to learn intricate relation-ships between pixels and object classes. This is particularly beneficial for challenging segmentation tasks. However, one significant challenge is the scarcity of labeled training data. To address this issue, self-supervised approaches using generative mod-els like Generative Adversarial Networks (GANs) provide a solution. GANs can gener-ate synthetic training data, which is useful for tasks where acquiring labeled training data is difficult or expensive, such as in medical imaging or remote sensing applica-tions. Additionally, GANs can generate realistic segmentation masks, which are cru-cial for tasks like medical imaging. In this study, we conducted a comparative analy-sis of DL-based segmentation approaches for COVID-19 CT scans. The evaluated approaches were assessed using metrics such as Dice Score, Specificity, and Sensitivi-ty. These metrics provide quantitative measures of the segmentation performance, allowing for an objective evaluation and comparison of the different techniques.

Multiple-Baseline Design Study of Deliberate Practice for Learning Psychotherapeutic Techniques

ABSTRACT Deliberate practice is a framework for expertise development that has been applied in many fields but rarely to studies of psychotherapy skill development. The current multiple-baseline study sought to examine the effects of deliberate practice on initial psychotherapy skill acquisition in three completely novice students. In addition to quantitative measures of performance, qualitative feedback concerning their methods of orienting to tasks was also collected regularly. Participants were given a minimal set of instructions to conduct and record psychotherapy role-play exercises, two of whom were provided with opportunities for focal deliberate practice beginning after their 3rd or 5th sessions. Results indicated that deliberate practice was associated with rapid, precipitous increases in reliably coded, structured evaluation of skill, which in many cases approached the cutoff for competent professional CBT practice. The potential of these methods to enhance learning and establish lifelong methods of skill improvement at all performance levels is discussed.

Abstract TMP11: Cerebral Visual Impairment in Pediatric Stroke Patients Assessed by Virtual Reality Testing

Introduction: Pediatric stroke is one of the leading causes of disability in children, many of whom incur brain injury affecting critical visual anatomical networks. Dedicated ophthalmologic examinations that detect cerebral visual impairment (CVI) - the leading cause of congenital visual impairment in developed countries - are not standard of care and can delay diagnosis and interventions. There is limited CVI data in pediatric stroke patients, and in this work we present data defining visual processing deficits in pediatric stroke patients. Hypothesis: We hypothesize that pediatric stroke patients will (1) demonstrate greater visual-spatial processing impairments than controls and (2) demonstrate increasing performance impairment as a function of increasing visual task demands. Methods: Using a virtual reality (VR) based visual search task system, we characterized significant deficits in higher-order visual processing of pediatric stroke patients (n = 17) as compared to healthy controls (n = 13). Stroke patients suffered from a perinatal (n = 8) or pediatric (n = 9) stroke - confirmed with brain MRI - affecting various vascular territories of the anterior and posterior circulations. The task required searching for on-screen targets with increasing visual complexity. Metrics such as reaction time, gaze error, shifted ellipse area, success rate, and pursuit saccade ratio were analyzed using unpaired group-level t-tests and considered significant if p &lt; 0.05. Results: Stroke patients took longer to identify the target (p = 0.001), were less accurate in finding it (p = 0.007), and exhibited greater impairment in performance with an increasing number of distractors compared to healthy children. Groups had comparable task compliance, as indexed by an off-screen count metric. Conclusion: This original work demonstrates that pediatric stroke patients, particularly those with strokes affecting the middle cerebral artery and posterior circulation vascular territories, exhibit impaired higher-order visual processing. This highlights the urgent need for standardized clinical testing to provide quantitative performance measures of cortical visual processing. This will allow timely detection and implementation of interventional therapies.

Citation metrics and strategic mutations of scientific research: narratives and evidence

Following the diffusion of the management model promoted by the New Public Management also within universities, sectors until then endowed with particular protection, such as the academic one, were progressively subjected to new controls and constraints, which gradually took the form of quantitative surveys, with a growing role of citation metrics. These evaluation processes have given rise to various important critical positions at an international level. If it is true that, according to Goodhart's law, "when a measure becomes an objective it ceases to be a good measure”, it is even more significant to note that when a measure becomes an objective, what is measured ceases to be what it was before. The quantitative measurement of academic performance has in fact triggered forms of gaming such as to alter the very game of scientific research, its purposes and the forms of its sharing. In the article we will carry out a comparison between the forms of mutation and the narratives that accompany them to see to what extent we are legitimate today to talk about the transformation of scientific research into strategic scientific research.

The Overall Collection Efficiency of Catching‐Type Precipitation Gauges in Windy Conditions

AbstractThe wind‐induced bias of catching‐type precipitation measurement instruments is quantified using Computational Fluid Dynamics with embedded liquid (raindrops) and solid (snowflakes) particle tracking. The performance of six common commercial gauges having different outer geometry is compared under a range of Precipitation Intensity (PI) and wind speed conditions by means of the numerically calculated catch ratios. Validation of the simulated aerodynamic behavior and its effect on water drop trajectories is provided by previous wind tunnel experiments. The functional dependence of the overall collection efficiency on the PI is derived as a quantitative measure of the instrument performance under a wind climatology with a uniform probability density function. Instruments with aerodynamic design produce a less pronounced diversion of hydrometeor trajectories—at any given size—than those having more traditional geometry. For liquid precipitation measurements, chimney‐shaped instruments rank low, while inverted conical and Nipher shielded instruments are quite performant solutions. The investigated quasi‐cylindrical gauges have intermediate behavior depending on their detailed aerodynamic features. For solid precipitation, all instruments rank low at light to moderate PI, except the Nipher shielded gauge. Results allow selecting the appropriate instrument for the local precipitation climatology at the measurement site and can be used to apply suitable adjustments to the measured precipitation.

Enhancing Product Excellence and Business Growth Approaches for Small and Medium-Sizes Pastry and Bakery Enterprises

TThis research aims to comprehensively assess the production facilities, production performance, operating income, the marketing mix, supporting services, and strategy development within Small and Medium-Sized Enterprises (SMEs) in the pastry and bakery industry. The study includes 41 respondents, comprising business owners, seven employees, 30 consumers, and three expert chefs in the pastry and bakery field. A mixed-method approach, combining descriptive qualitative and quantitative analyses, was employed. The research involved both qualitative exploration of respondents' perceptions and quantitative measurements of production performance and marketing effectiveness. Findings revealed satisfactory product quality, yielding profitable results from an average daily production of 100 bread loaves and 32 cakes. While the overall marketing mix was effective, the promotional aspect was identified as an area requiring enhancement. Supporting services were underutilized, suggesting opportunities for improvement, particularly in optimizing raw material utilization to heighten productivity and income while ensuring superior product quality. The study underscores the need for strategic enhancements in promotional efforts and the maximization of supporting services to align with the already satisfactory product quality. Furthermore, it emphasizes the significance of leveraging raw materials effectively to augment productivity and profitability in the SMEs pastry and bakery sector. Enhancing promotional strategies and optimizing supporting services are crucial for SMEs in the pastry and bakery industry to strengthen market positioning. Additionally, future research endeavors should delve deeper into the impact of optimized raw material utilization on productivity and income within this sector, thereby fostering sustained growth and competitiveness.

Mathematical modeling of human memory.

The mathematical study of human memory is still an open challenge. Cognitive psychology and neuroscience have given a big contribution to understand how the human memory is structured and works. Cognitive psychologists developed experimental paradigms, conceived quantitative measures of performance in memory tasks for both healthy people and patients with memory disorders, but in terms of mathematical modeling human memory there is still a lot to do. There are many ways to mathematically model human memory, for example, by using mathematical analysis, linear algebra, statistics, and artificial neural networks. The aim of this study is to provide the reader with a description of some prominent models, involving mathematical analysis and linear algebra, designed to describe how memory works by predicting the results of psychological experiments. We have ordered the models from a chronological point of view and, for each model, we have emphasized what are, in our opinion, the strong and weak points. We are aware that this study covers just a part of human memory modeling as well as that we have made a personal selection, which is arguable. Nevertheless, our hope is to help scientists to modeling human memory and its diseases.

Deep Learning and Gastric Cancer: Systematic Review of AI-Assisted Endoscopy.

Gastric cancer (GC), a significant health burden worldwide, is typically diagnosed in the advanced stages due to its non-specific symptoms and complex morphological features. Deep learning (DL) has shown potential for improving and standardizing early GC detection. This systematic review aims to evaluate the current status of DL in pre-malignant, early-stage, and gastric neoplasia analysis. A comprehensive literature search was conducted in PubMed/MEDLINE for original studies implementing DL algorithms for gastric neoplasia detection using endoscopic images. We adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The focus was on studies providing quantitative diagnostic performance measures and those comparing AI performance with human endoscopists. Our review encompasses 42 studies that utilize a variety of DL techniques. The findings demonstrate the utility of DL in GC classification, detection, tumor invasion depth assessment, cancer margin delineation, lesion segmentation, and detection of early-stage and pre-malignant lesions. Notably, DL models frequently matched or outperformed human endoscopists in diagnostic accuracy. However, heterogeneity in DL algorithms, imaging techniques, and study designs precluded a definitive conclusion about the best algorithmic approach. The promise of artificial intelligence in improving and standardizing gastric neoplasia detection, diagnosis, and segmentation is significant. This review is limited by predominantly single-center studies and undisclosed datasets used in AI training, impacting generalizability and demographic representation. Further, retrospective algorithm training may not reflect actual clinical performance, and a lack of model details hinders replication efforts. More research is needed to substantiate these findings, including larger-scale multi-center studies, prospective clinical trials, and comprehensive technical reporting of DL algorithms and datasets, particularly regarding the heterogeneity in DL algorithms and study designs.

Novel ensemble feature selection techniques applied to high-grade gastroenteropancreatic neuroendocrine neoplasms for the prediction of survival

Background and ObjectiveDetermining the most informative features for predicting the overall survival of patients diagnosed with high-grade gastroenteropancreatic neuroendocrine neoplasms is crucial to improve individual treatment plans for patients, as well as the biological understanding of the disease. The main objective of this study is to evaluate the use of modern ensemble feature selection techniques for this purpose with respect to (a) quantitative performance measures such as predictive performance, (b) clinical interpretability, and (c) the effect of integrating prior expert knowledge. MethodsThe Repeated Elastic Net Technique for Feature Selection (RENT) and the User-Guided Bayesian Framework for Feature Selection (UBayFS) are recently developed ensemble feature selectors investigated in this work. Both allow the user to identify informative features in datasets with low sample sizes and focus on model interpretability. While RENT is purely data-driven, UBayFS can integrate expert knowledge a priori in the feature selection process. In this work, we compare both feature selectors on a dataset comprising 63 patients and 110 features from multiple sources, including baseline patient characteristics, baseline blood values, tumor histology, imaging, and treatment information. ResultsOur experiments involve data-driven and expert-driven setups, as well as combinations of both. In a five-fold cross-validated experiment without expert knowledge, our results demonstrate that both feature selectors allow accurate predictions: A reduction from 110 to approximately 20 features (around 82%) delivers near-optimal predictive performances with minor variations according to the choice of the feature selector, the predictive model, and the fold. Thereafter, we use findings from clinical literature as a source of expert knowledge. In addition, expert knowledge has a stabilizing effect on the feature set (an increase in stability of approximately 40%), while the impact on predictive performance is limited. ConclusionsThe features WHO Performance Status, Albumin, Platelets, Ki-67, Tumor Morphology, Total MTV, Total TLG, and SUVmax are the most stable and predictive features in our study. Overall, this study demonstrated the practical value of feature selection in medical applications not only to improve quantitative performance but also to deliver potentially new insights to experts.