Quality Prediction Research Articles

Physics-Informed Neural ODE with Heterogeneous Control Inputs (PINOHI) for Quality Prediction of Composite Adhesive Joints

Composite materials have long been used in various industries due to their superior properties such as high strength, lightweight and corrosive resistance. Bonded composite joints are finding increasing applications as they provide extensive structural benefits and design flexibility. On the other hand, the failure mechanism of composite adhesive joints is not fully understood. A model that bridges manufacturing parameters and final quality measures is highly desired for the design and optimization of the composite adhesive joints manufacturing process. In this study, a novel framework of physics-informed Neural Ordinary Differential Equation (ODE) with heterogeneous control input (PINOHI) is proposed, which links the heterogeneous manufacturing parameters to the final bonding quality of composite joints. The proposed model structure is heavily motivated by engineering knowledge, by incorporating a calibrated mathematical physics model into the Neural ODE framework, which can significantly reduce the number of data samples required from costly experiments while maintaining high prediction accuracy. The proposed PINOHI model is implemented in the quality prediction of composite adhesive joints bonding problem. A set of experiments and associated data analytics are conducted to demonstrate the superior property of the PINOHI model by using both the leave-one-batch-out cross-validation and sensitivity analysis.

Short-term prediction of horizontal winds in the mesosphere and lower thermosphere over coastal Peru using a hybrid model

The mesosphere and lower thermosphere (MLT) are transitional regions between the lower and upper atmosphere. The MLT dynamics can be investigated using wind measurements conducted with meteor radars. Predicting MLT winds could help forecast ionospheric parameters, which has many implications for global communications and geo-location applications. Several literature sources have developed and compared predictive models for wind speed estimation. However, in recent years, hybrid models have been developed that significantly improve the accuracy of the estimates. These integrate time series decomposition and machine learning techniques to achieve more accurate short-term predictions. This research evaluates a hybrid model that is capable of making a short-term prediction of the horizontal winds between 80 and 95 km altitudes on the coast of Peru at two locations: Lima (12°S, 77°W) and Piura (5°S, 80°W). The model takes a window of 56 data points as input (corresponding to 7 days) and predicts 16 data points as output (corresponding to 2 days). First, the missing data problem was analyzed using the Expectation Maximization algorithm (EM). Then, variational mode decomposition (VMD) separates the components that dominate the winds. Each resulting component is processed separately in a Long short-term memory (LSTM) neural network whose hyperparameters were optimized using the Optuna tool. Then, the final prediction is the sum of the predicted components. The efficiency of the hybrid model is evaluated at different altitudes using the root mean square error (RMSE) and Spearman’s correlation (r). The hybrid model performed better compared to two other models: the persistence model and the dominant harmonics model. The RMSE ranged from 10.79 to 27.04 m s−1, and the correlation ranged from 0.55 to 0.94. In addition, it is observed that the prediction quality decreases as the prediction time increases. The RMSE at the first step reached 6.04 m s−1 with a correlation of 0.99, while at the sixteenth step, the RMSE increased up to 30.84 m s−1 with a correlation of 0.5.

Improving prediction of groundwater quality in situations of limited monitoring data based on virtual sample generation and Gaussian process regression

The increasing pollution of aquifers by human activities over recent decades poses a threat to drinking water safety. While Gaussian Process Regression (GPR) is a robust tool for predicting and monitoring water quality, its effectiveness is hindered limitations of available data on model training and validation, known as the “small sample problem”. Various attempts to resolve this problem include virtual sample generation (VSG). This study aimed to increase the accuracy of GPR for predicting water quality in situations of limited datasets. Three VSG methods, namely Multi Distribution Mega-Trend Diffusion (MD-MTD), Generative Adversarial Network (GAN), and t-distributed stochastic nearest neighbor embedding (t-SNE) were compared for enhancing the accuracy of GPR model prediction of Strontium (Sr2+). The models were used to predict Sr2+ in the shallow aquifer system in Songyuan, Jilin Province. The results showed that t-SNE provided the most significant improvement to the accuracy of the GPR, with R2 increasing from 0.86 to 0.99 (12.98 %), followed by MD-MTD (R2 of 0.95, 9.39 %), with the least improvement obtained by GAN (R2 of 0.92, 5.98 %). Boxplots show that MD-MTD-GPR predictions do not fully capture observed data distributions. GANs accurately replicate the data distribution, while t-SNE-GPR achieves the highest prediction accuracy and handles data fluctuations. GPR accuracy improves with an increasing number of virtual samples but tends to decrease when the number exceeds 258 in this study. This study can guide the improvement of the accuracy of GPR for situations of limited datasets. The results of this study can help improve water quality management and drinking water safety in regions with sparse monitoring data.

Predicting image quality of forensic footwear impressions

The ability of a footwear examiner to confidently discern features of importance in a forensic examination is directly related to impression quality. As a result, quality directly impacts the strength an examiner can ascribe to any opinion of source attribution. Despite the importance of image quality during both the analysis and comparison phases of an examination, there is limited research on the estimation, variation, and prediction of footwear impression quality. In response, this study aims to develop a methodology for evaluating footwear impression quality by regressing image features against subjective judgments of quality. Using a dataset of more than 450 impressions evaluated by more than 40 participants, estimates of intra- and inter-rater consistency were computed. After identifying reliable raters, matrix completion was performed, thereby permitting data imputation. This was based on an approximate 90:10 split between training and testing data, ultimately resulting in 6,000 quality predictions, which exhibited a mean agreement with ground truth > 95%. Each image in the dataset was then reduced to a 10-dimensional feature vector describing image attributes such as complexity, contrast, sharpness, and noise. These features were used to train multiple ordinal and multinomial regression models aimed at predicting image quality. Model comparison resulted in an overall optimal accuracy of approximately 75% when comparing predicted/modeled quality against the ground truth of subjective rater opinions. The resulting semi-automated, reference-free and numerical prediction tool exhibits reasonable success for impression quality prediction when presented with the types of challenging and domain-specific images often encountered in forensic footwear comparisons. The outcomes and challenges associated with this investigation provide a foundation upon which future studies of quality can be modified and built, with the downstream goal of an increased understanding of how quality impacts weight of evidence.

Exploring Air Quality Dynamics and Predictive Modeling by Using Artificial Intelligence During COVID-19 Lock Down Over the Western Part of India

The lockdown period, initially imposed for three months due to the COVID-19 outbreak in India, was later prolonged. Air quality data from eight monitoring sites in Rajasthan was used to calculate the AQI according to the following parameters: Particulate matter (PM2.5 and PM10), Nitrogen Dioxide (NO2), Ammonia (NH3), Sulfur dioxide (SO2), Ozone (O3), and Carbon monoxide (CO), dispersed throughout the state by CPCB. Among the chosen cities, the study found that the AQI percentage dropped the most in Alwar, by 35.6% between pre-lockdown and lockdown. Conversely, it rose the most in Jaipur, by 86.77% between lockdown and post-lockdown. Python deep learning was used to simulate the relationship between Air Quality Index and Air contamination in the study area. Air quality index values ranging from Good (0–50) to Severe (>401) were used to create the AQI class categorization in Python. The study found that PM2.5 and PM10 had the strongest correlation. Metrics such as the coefficient of determination (R2) and the root mean square error (RMSE) were applied to assess the model on the datasets used for training and testing. Random forest, decision trees, and linear regression were worked to verify the precision of the prototype. The author used supervised learning techniques, such as decision tree (DT), extreme gradient boosting (XGBoost), K-nearest neighbor (KNN), logistic regression (LR), and random forest (RF), to determine the model's prediction. These findings suggest that urban areas are characterized by societal, commercial, and cultural aspects that contribute to similar discharge patterns and air quality issues. The study would be advantageous for authorities, as it is clearly apparent that reducing the sources of emissions can improve quality. This will set the stage for safeguarding and improving the environment.

AI-based prediction of academic success: Support for many, disadvantage for some?

The use of computational tools to predict academic success has become increasingly popular. Machine learning algorithms, trained on past study histories, have been shown to provide valid predictions. However, knowing about biases and unfairness in algorithms, one should take a closer look at these predictions. This paper explores the extent to which the predictive accuracy of academic success varies between specific groups of students, focusing on traditional and non-traditional students (NTS), who have not acquired a higher education entrance qualification at school. In a case study the study compares several popular algorithms and their prediction quality, and investigates whether misclassified NTS show positive or negative biases. Results revealed that the accuracy of predicting academic success for NTS was significantly lower than when considering all students as a whole. The direction of the distortion cannot be determined exactly due to small case numbers. The study emphasizes that the possibility of bias always has to be considered when predicting study success, and the use of such tools must ensure there are no undesirable biases that could affect certain students.

The impact of smartphone dependence on college students' sleep quality: the chain-mediated role of negative emotions and health-promoting behaviors.

Sleep disturbances among college students have become a significant issue affecting their daily lives. This study aims to explore the relationship between smartphone dependence and sleep quality and examine the mediating roles of negative emotions and health-promoting behaviors. A total of 23,652 college students were included in the study, and 21,314 valid questionnaires were collected. The survey assessed demographic factors, smartphone dependence, sleep quality, negative emotions, and health-promoting behaviors. A chain mediation analysis was conducted to examine the relationships among these factors. Smartphone dependence was significantly positively correlated with sleep quality (r = 0.272, p < 0.001) and negative emotions (r = 0.414, p < 0.001), and significantly negatively correlated with health-promoting behaviors (r = -0.178, p < 0.001). Sleep quality was positively correlated with negative emotions (r = 0.472, p < 0.001) and negatively correlated with health-promoting behaviors (r = -0.218, p < 0.001).Smartphone dependence was a significant positive predictor of sleep quality. Moreover, negative emotions and health-promoting behaviors influenced the relationship between smartphone dependence and sleep quality. The total effect, direct effect, and indirect effect values were 0.304, 0.122, and 0.170, respectively. Different demographic factors (such as gender and place of residence) can lead to variations in different variables. Smartphone dependence and negative emotions have a positive impact on sleep quality among college students, while health-promoting behaviors have a negative impact. Smartphone dependence directly and positively affects sleep quality and can also influence it indirectly through the mediating effects of negative emotions and health-promoting behaviors, both individually and in a chain-like manner.

An Interval Neural Network Method for Identifying Static Concentrated Loads in a Population of Structures

During the design and validation of structural engineering, the focus is on a population of similar structures, not just one. These structures face uncertainties from external environments and internal configurations, causing variability in responses under the same load. Identifying the real load from these dispersed responses is a significant challenge. This paper proposes an interval neural network (INN) method for identifying static concentrated loads, where the network parameters are internalized to create a new INN architecture. Additionally, the paper introduces an improved interval prediction quality loss function indicator named coverage and mean square criterion (CMSC), which balances the interval coverage rate and interval width of the identified load, ensuring that the median of the recognition interval is closer to the real load. The efficiency of the proposed method is assessed through three examples and validated through comparative research against other loss functions. Our research findings indicate that this approach enhances the interval accuracy, robustness, and generalization of load identification. This improvement is evident even when faced with challenges such as limited training data and significant noise interference.

Psychiatric predictors of quality of life in Parkinson's disease: A three-year longitudinal study

IntroductionParkinson's disease (PD) is associated with worsened quality of life (QOL) over time. Few longitudinal studies exist investigating the relationship of psychiatric comorbidities with QOL in people with PD (PwP). We sought to determine specific psychiatric symptoms associated with decreasing QOL in PwP over time. MethodsWe recruited PwP without dementia from a movement disorders clinic at an academic medical center. Participants were evaluated annually with motor and neuropsychological assessments at each visit. QOL was measured using the Parkinson's Disease Questionnaire-39 (PDQ-39). We assessed psychiatric symptoms, including depression (Beck Depression Inventory II, BDI-II), anxiety (Beck Anxiety Index, BAI), and apathy (Apathy Scale). Psychosis and impulse control disorders (ICDs) were recorded as present or absent. Using random coefficient regression, we analyzed psychiatric features associated with worsened QOL in PwP over three years. ResultsFrom the 105 participants enrolled at baseline, 67 completed three years of follow up. Mean PDQ-39 scores increased from 16.0 at baseline to 19.8 at year three. In multivariate analysis, higher BDI-II scores, BAI scores, and apathy scores were uniquely associated with worsened QOL over time (p < 0.001 for all measures), while presence of ICDs (p = 0.18) or psychosis (p = 0.10) were not. Changes in the BAI score and the BDI-II score exerted similar effects on the overall PDQ-39 score. ConclusionDepression, anxiety, and apathy are all associated with worsening quality of life over time in PwP, while presence of ICDs and psychosis are not. Treatment of these symptoms may lead to improved QOL in PwP.

Research on Machining Quality Prediction Method Based on Machining Error Transfer Network and Grey Neural Network

Machining quality prediction is the critical link of quality control in parts machining. With the advent of the Industry 4.0 era, intelligent manufacturing and data-driven technologies bring new ideas for quality control in complex machining processes. Quality control is complicated for multi-process, multi-condition, small-batch, and high-precision parts processing requirements. To solve this problem, this paper proposes a machining quality prediction method based on the machining error transfer network and the grey neural network. Initially, by constructing a processing error transfer network, the error transfer law in part processing is described, and the PageRank algorithm and the influence degree of the nodes are used to determine the critical quality features. Additionally, the problem of low prediction accuracy due to small sample data and multiple coupling relationships is solved using the grey neural network algorithm, and a high accuracy prediction of critical quality features is achieved. Finally, the effectiveness and reliability of the method are verified by the case of medium-speed marine diesel engine fuselage processing. The results indicate that this method not only effectively identifies critical quality features in the machining process of complex parts, but it also maintains a high predictive accuracy for these features, even with small samples and limited data.

A Scoping Review of the Predictive Qualities of Walking Speed in Older Adults.

Walking speed (WS) is an easily assessable and interpretable functional outcome measure with great utility for the physical therapist providing care to older adults. Since WS was proposed as the sixth vital sign, research into its interpretation and use has flourished. The purpose of this scoping review is to identify the current prognostic value of WS for the older adult. A scoping review was conducted using PubMed, CINAHL, and SPORTDiscus to find relevant articles highlighting the predictive capabilities of WS for older adults. Titles and abstracts were reviewed to identify relevant articles. Articles were excluded based on the following criteria: sample included both younger and older adults without separate analyses, sample was focused on a particular disease, if the study was published before 2017, or if the study did not report relevant cut points for interpretation of WS. The search returned 1064 results. Following removal of articles not meeting inclusion criteria and critical appraisal, relevant cut points were extracted from 47 original research publications. A preliminary review of the included articles showed that WS is a valuable prognostic tool across many health domains, including mental health, mortality, disability, pain, bone and joint health, falls, cognition, physical activity, metabolic health, risk for cardiovascular disease, socialization, and metabolic health. The fastest WS of 1.32 meters per second (m/s) served as a cutoff for decreased risk for incident development of type 2 diabetes, while the slowest WS of less than 0.2 m/s was associated with increased duration of hospitalization. Multiple studies reported on the prognostic value of WS slower than 1.0 m/s. Although the reported range of predictive WS values was broad, multiple studies found WS of approximately 1.0 m/s to be a useful marker for delineating risk or decline across a variety of health domains. Clinicians may find it useful to use a WS slower than 1.0 m/s as a "yellow flag" to guide evaluation and intervention for their older adult clients.

Multiphase time-varying batch process monitoring based on stage moving model migration

To improve product qualities and obtain maximum comprehensive economic benefits from batch process production, it is of great practical significance to optimize the operation of batch production processes. Considering the multiphase and slow time-varying characteristics, a multiphase time-varying batch process monitoring method is proposed based on stage moving model migration technology. Firstly, the sliding windows are constructed with moving phases, where the moving of different phases is decided by the influence of the process variables to the final product. And time slices are the basic units for building the regression models. Secondly, to analyze and capture the relationship between the process variables of both the old mode and new mode and the final quality variable, the JY-PLS algorithm is utilized on each time slice within sliding windows for quality prediction. Then, monitoring indexes and corresponding control limits are calculated and used for process monitoring. The application to a real multiphase time-varying batch process, the start-up process of injection molding process, proved the effectiveness of the proposed method.

Analysis of the Impact of Meteorological Factors on Predicting Air Quality in South Tangerang City using Random Forest Method

Air pollution has become one of the most significant environmental problems in many cities throughout the world, which can endanger public health and the environment. Understanding the impact of meteorological conditions on air quality is very important to understanding air pollution patterns. This study investigates the influence of meteorological variables on air quality predictions in South Tangerang City, Indonesia, using the Random Forest method. Modeling is carried out by building two scenarios, namely predictions using meteorological variables and predictions without meteorological variables. Prediction performance analysis is measured using MAE, MSE, RMSE, R-square, and accuracy. The accuracy results of the research show that predictions without meteorological variables provide good prediction results with a value of 86.42%, but predictions with meteorological variables have better performance with a value reaching 98.99%. The largest error values from each model were 2.58 MAE, 71.82 MSE, and 8.4747 RMSE obtained in prediction modeling without meteorological variables, while the smallest error values were obtained in prediction modeling using meteorological variables, namely 0.00, 0.01, and 0.0219, respectively, for MAE, MSE, and RMSE. This research contributes to a better understanding of the relationship between meteorology and air pollution and air quality in urban areas and helps develop targeted mitigation strategies to improve air quality and public health, especially in South Tangerang City and the surrounding area.

Comparative analysis of neural network architectures for short-term FOREX forecasting

This document outlines the analysis, design, implementation, and benchmarking of various neural network architectures in a short-term frequency prediction system for the FOREX market. Our objective is to emulate the decision-making process of a human expert (technical analyst) through a system that swiftly adapts to market condition changes, thereby optimizing short-term trading strategies. We have designed and implemented a series of LSTM neural network architectures that take exchange rate values as input to generate short-term market trend forecasts. Additionally, we developed a custom ANN architecture based on simulators for technical analysis indicators. We performed a comparative analysis of the results and came to useful conclusions regarding the suitability of each architecture and the cost in terms of time and computational power to implement them. The ANN custom architecture produces better prediction quality with higher sensitivity using fewer resources and spending less time than LSTM architectures. The ANN custom architecture appears to be ideal for use in low-power computing systems and for use cases that need fast decisions with the least possible computational cost.

Sleep Quality, Daytime Sleepiness and Sleep Hygiene Among Youth in a Rural District in South India: A Cross-Sectional Study.

Background Adequate sleep is crucial for youth cognitive function, academic performance, and mental health. However, various factors, including academic pressure, technology use, and socio-cultural norms can significantly impact sleep patterns, particularly in rural settings. This cross-sectional study assessed sleep quality, daytime sleepiness prevalence, and sleep hygiene practices among youth in a rural South Indian district. We also investigated factors associated with these sleep parameters in this understudied population Methods This was a cross-sectional study among 852 young subjects who were assessed with a self-reported proforma of socio-demographic details, behavioural factors,Pittsburgh Sleep Quality Index and Epworth Sleepiness Scale. Regression model was created to analyse the predictor for sleep quality and daytime sleepiness with respect to sociodemographic variables, behavioural factors and sleep hygiene practices. Results Our study revealed that 49.4% (n=421) of participants exhibited poor sleep quality, while 29.5% (n=251) reported abnormal daytime sleepiness. The most prevalent sleep hygiene practices were reading in bed (68.5%, n=584) and pre-bedtime eating (56.22%, n=479). Multivariate analysis indicated that sleep quality was significantly associated with accommodation type, with increased odds for those in private accommodation (adjusted odds ratio (AOR) 1.29, 95% CI: 1.14-1.62) and hostels (AOR 3.79, 95% CI: 1.79-8.06). Additionally, eating in bed (AOR 1.42, 95% CI: 1.03-1.95) and pre-bedtime eating (AOR 1.41, 95% CI: 1.07-1.88) were also associated with poor sleep quality. Factors significantly associated with daytime sleepiness included younger age (AOR 0.84, 95% CI: 0.75-0.94), non-medical academic streams (AOR 1.94, 95% CI: 1.33-2.83), extensive internet usage (three or more hours) (AOR 1.87, 95% CI: 1.15-3.13), watching TV in bed (AOR 1.46, 95% CI: 1.06-1.99), writing (AOR 1.45, 95% CI: 1.02-2.06), and eating in bed (AOR 1.55, 95% CI: 1.09-2.21). Conclusion This study reveals a significant incidence of poor sleep quality and daytime drowsiness among young individuals residing in rural areas of South India. The results of our study emphasize significant connections between sleep disturbances and several modifiable aspects, such as the kind of accommodation, eating habits, and use of technology. The impact of eating behaviours, both in bed and before bedtime, on sleep quality and daytime sleepiness underscores the importance of proper sleep hygiene education. Furthermore, the relationship between extensive internet usage and daytime sleepiness points to the growing influence of digital technology on youth sleep patterns. These findings emphasize the need for comprehensive sleep health programs tailored to rural youth. Such initiatives should address environmental factors, promote healthy sleep hygiene practices, and provide guidance on balanced technology use. Additionally, the varying impact of academic streams on sleep parameters suggests that sleep health strategies may need to be customized for different educational contexts.

Waste management and water quality evaluation prediction in urban environments through advanced robust hybrid machine learning algorithms

Water quality management is a crucial aspect of environmental protection, requiring the monitoring and regulation of effluent discharges into surface water bodies. This research introduces a novel approach to predicting Water Quality Evaluation (WQE) through a unique hybrid model, ABC-DWKNN-ICA, which integrates the Distance-weighted K-Nearest Neighbors (DWKNN) algorithm with the Artificial Bee Colony (ABC), Firefly Algorithm (FA), Imperialist Competitive Algorithm (ICA), and Gravitational Search Algorithm (GSA). Utilizing a comprehensive dataset of 1106 data points from Telangana, India, spanning 2018–2020, the study examines a range of water quality parameters, including Ground Water Level (GWL), Potential of Hydrogen (PH), Electrical Conductivity (EC), and others. The ABC-DWKNN-ICA model demonstrates exceptional performance in terms of Recall, Precision, Accuracy, and F1 Score for WQE prediction, distinguishing itself with enhanced feature selection, improved classification accuracy, robustness to noise and outliers, reduced dimensionality, and scalability to large datasets. This hybrid model represents a significant advancement over existing approaches, including traditional Hybrid Machine Learning (HML) algorithms such as ABC-DWKNN, FA-DWKNN, ICA-DWKNN, and GSA-DWKNN. By focusing on the capabilities of ABC-DWKNN-ICA rather than comparing all HML algorithms, the research highlights its superior effectiveness in water quality prediction, with performance metrics of 0.83 for Recall, 0.86 for Precision, 0.91 for Accuracy, and 0.86 for F1 Score. This study thus fills a critical research gap by demonstrating the model's value in environmental data analysis and offering promising prospects for more effective management of water resources. Additionally, feature selection, dimensionality reduction, enhanced accuracy, noise handling, and imbalanced dataset management are key advantages of the proposed model.

Research on Critical Quality Feature Recognition and Quality Prediction Method of Machining Based on Information Entropy and XGBoost Hyperparameter Optimization

To address the problem of predicting machining quality for critical features in the manufacturing process of mechanical products, a method that combines information entropy and XGBoost (version 2.1.1) hyperparameter optimization is proposed. Initially, machining data of mechanical products are analyzed based on information entropy theory to identify critical quality characteristics. Subsequently, a quality prediction model for these critical features is established using the XGBoost machine learning framework. The model’s hyperparameters are then optimized through Bayesian optimization. This method is applied as a case study to a medium-speed marine diesel engine piston. After the critical quality characteristics in the machining process are identified, the machining quality of these vital characteristics is predicted, and the results are compared with those obtained from a machine learning model without hyperparameter optimization. The findings demonstrate that the proposed method effectively predicts the machining quality of mechanical products.

Advanced manufacturing with machine learning: enhancing predictive maintenance, quality control, and process optimization

Advanced manufacturing with machine learning: enhancing predictive maintenance, quality control, and process optimization

A modified parametric model to predict visco-elastic properties of magneto-rheological elastomers at non-LVE region

The magneto-rheological elastomer is mostly used in vibration isolation, for which higher modulus and lower Payne effect factors are crucial parameters. Strain amplitudes and frequencies influence many applications under dynamic modes. In this work, the dynamic viscoelastic properties of MRE, fabricated using electrolyte iron (EI) particles, were measured for varying strain amplitude, magnetic field and frequency. A fractional Kelvin-Voigt (KV) model is used in a frequency region from 0.01 to 40 Hz to predict the rheological behaviour. However, the available models failed to explain the observed behaviour at low frequencies and high magnetic fields and increasing strain amplitude (i.e. in the non-viscoelastic region). Therefore, a new modified KV model is proposed in this work to incorporate the drawbacks and hence can validate for varying frequency, magnetic field and strain amplitudes. The added terms can also be used in the fractional derivative Maxwell model to explain the effect of strain amplitude and magnetic field at various frequencies. The proposed model significantly improves the quality of experimental prediction in the low-frequency range, corresponding to a slow dissipative process at different strain amplitudes.

Prediction of tumor board procedural recommendations using large language models.

Multidisciplinary tumor boards are meetings where a team of medical specialists, including medical oncologists, radiation oncologists, radiologists, surgeons, and pathologists, collaborate to determine the best treatment plan for cancer patients. While decision-making in this context is logistically and cost-intensive, it has a significant positive effect on overall cancer survival. METHODS: We evaluated the quality and accuracy of predictions by several large language models for recommending procedures by a Head and Neck Oncology tumor board, which we adapted for the task using parameter-efficient fine-tuning or in-context learning. Records were divided into two sets: n=229 used for training and n=100 records for validation of our approaches. Randomized, blinded, manual human expert classification was used to evaluate the different models. RESULTS: Treatment line congruence varied depending on the model, reaching up to 86%, with medically justifiable recommendations up to 98%. Parameter-efficient fine-tuning yielded better outcomes than in-context learning, and larger/commercial models tend to perform better. Providing precise, medically justifiable procedural recommendations for complex oncology patients is feasible. Extending the data corpus to a larger patient cohort and incorporating the latest guidelines, assuming the model can handle sufficient context length, could result in more factual and guideline-aligned responses and is anticipated to enhance model performance. We, therefore, encourage further research in this direction to improve the efficacy and reliability of large language models as support in medical decision-making processes.