Tag: statistics

BMC Med Inform Decis Mak. 2025 Mar 3;25(1):105. doi: 10.1186/s12911-025-02937-5.

ABSTRACT

Current research on the association between demographic variables and dietary patterns with atherosclerotic cardiovascular disease (ASCVD) is limited in breadth and depth. This study aimed to construct a machine learning (ML) algorithm that can accurately and transparently establish correlations between demographic variables, dietary habits, and ASCVD. The dataset used in this research originates from the United States National Health and Nutrition Examination Survey (U.S. NHANES) spanning 1999-2018. Five ML models were developed to predict ASCVD, and the best-performing model was selected for further analysis. The study included 40,298 participants. Using 20 population characteristics, the eXtreme Gradient Boosting (XGBoost) model demonstrated high performance, achieving an area under the curve value of 0.8143 and an accuracy of 88.4%. The model showed a positive correlation between male sex and ASCVD risk, while age and smoking also exhibited positive associations with ASCVD risk. Dairy product intake exhibited a negative correlation, while a lower intake of refined grains did not reduce the risk of ASCVD. Additionally, the poverty income ratio and calorie intake exhibited non-linear associations with the disease. The XGBoost model demonstrated significant efficacy, and precision in determining the relationship between the demographic characteristics and dietary intake of participants in the U.S. NHANES 1999-2018 dataset and ASCVD.

PMID:40033349 | DOI:10.1186/s12911-025-02937-5

Scand J Trauma Resusc Emerg Med. 2025 Mar 3;33(1):35. doi: 10.1186/s13049-025-01350-1.

ABSTRACT

BACKGROUND: For preoxygenation, German guidelines consider non-invasive ventilation (NIV) as a possible method in prehospital trauma care in the absence of aspiration, severe head or face injuries, unconsciousness, or patient non-compliance. As data on the utilization and characteristics of patients receiving NIV are lacking, this study aims to identify predictors of NIV usage in trauma patients using machine learning and compare these findings with the current national guideline.

METHODS: A cross-regional registry of prehospital emergency services in southwestern Germany was searched for cases of emergency anesthesia in multiply injured patients in the period from 2018 to 2020. Initial vital signs, oxygen saturation, respiratory rate, heart rate, systolic blood pressure, Glasgow Coma Scale (GCS), injury pattern, shock index and age were examined using logistic regression. A decision tree algorithm was then applied in parallel to reduce the number of attributes, which were subsequently tested in several machine learning algorithms to predict the usage of NIV before the induction of anesthesia.

RESULTS: Of 992 patients with emergency anesthesia, 333 received NIV (34%). Attributes with a statistically significant influence (p < 0.05) in favour of NIV were bronchial spasm (odds ratio (OR) 119.75), dyspnea/cyanosis (OR 2.28), moderate and severe head injury (both OR 3.37) and the respiratory rate (OR 1.07). Main splitting points in the initial decision tree included auscultation (rhonchus and bronchial spasm), respiratory rate, heart rate, age, oxygen saturation and head injury with moderate head injury being more frequent in the NIV group (23% vs. 12%, p < 0.01). The rates of aspiration and the level of consciousness were equal in both groups (0.01% and median GCS 15, both p > 0.05). The prediction accuracy for NIV usage was high for all algorithms, except for multilayer perceptron and logistic regression. For instance, a Bayes Network yielded an AUC-ROC of 0.96 (95% CI, 0.95-0.96) and PRC-areas of 0.96 [0.96-0.96] for predicting and 0.95 [0.95-0.96] for excluding NIV usage.

CONCLUSIONS: Machine learning demonstrated an excellent categorizability of the cohort using only a few selected attributes. Injured patients without severe head injury who presented with dyspnea, cyanosis, or bronchial spasm were regularly preoxygenated with NIV, indicating a common prehospital practice. This usage appears to be in accordance with current German clinical guidelines. Further research should focus on other aspects of the decision making like airway anatomy and investigate the impact of preoxygenation with NIV in prehospital trauma care on relevant outcome parameters, as the current evidence level is limited.

PMID:40033329 | DOI:10.1186/s13049-025-01350-1

BMC Sports Sci Med Rehabil. 2025 Mar 3;17(1):34. doi: 10.1186/s13102-025-01088-4.

ABSTRACT

Performance analysis can provide coaches with a range of relevant information and support more informed decision-making. The objective of this research was to determine running performance (RP) within five-minute intervals when scoring and conceding goals in the UEFA Champions League (UCL). Matches from the UCL 2020/2021 season were analyzed, and relevant data were retrieved using the InStat Fitness semi-automatic video system. Statistical analysis employed one-way analysis of variance (ANOVA) for comparisons and partial eta squared (η2) to determine effect size. Team performance was determined by measuring total distance covered (TD) and high-intensity running (HIR) when the team scored a goal, conceded a goal, and when the score did not change. Our primary results indicated significant differences in three out of 20 five-minute intervals for the TD parameter and four out of 20 for HIR when teams scored goals. There were also significant differences in eight out of 20 intervals for TD and three out of 20 for HIR when teams conceded goals. In conclusion, significant goal concessions were observed during all the five-minute intervals in which teams substantially reduced their RP. From a practical point of view, coaches should be aware, especially in the context of the pacing strategy used, that team RP affects the scoreline directly and the match outcome indirectly.

PMID:40033321 | DOI:10.1186/s13102-025-01088-4

BMC Public Health. 2025 Mar 3;25(1):838. doi: 10.1186/s12889-025-22040-8.

ABSTRACT

BACKGROUND: Evidence on psychological outcomes of gaming disorder (GD) is still scarce. This study aimed to investigate the mediating role of sleep hygiene in the relationship between GD and psychological distress (depression and anxiety) among Iranian adolescents.

METHODS: This was a cross-sectional study among school students in Qazvin city, Iran. We administered GD, anxiety, and depression questionnaires in a paper-and-pencil format. GD was measured using the GD S4-SF scale, and anxiety and depression were evaluated using the DASS-21. We assessed sleep health as a mediator using the Sleep Hygiene Behaviors scale. Covariance-Based Structural Equation Modeling (CB-SEM) was employed for data analysis, accounting for sex and physical activity as the main confounders. Statistical significance was determined using various fit indices and confidence intervals.

RESULTS: The sample consisted of 600 adolescents (41% female). CB-SEM revealed a positive but not statistically significant association between GD and depression, along with a negative statistically significant association with anxiety. Notably, sleep hygiene was identified as a partial mediator in the relationship between GD and depression, indicating that poor sleep practices may exacerbate depressive symptoms among adolescents with GD. However, no mediating effect was observed for anxiety.

CONCLUSION: Our data supported a mediating role for sleep hygiene in the association between GD and depression among participants. Our results highlight the critical need for targeted policy interventions to improve sleep hygiene among adolescents with GD.

PMID:40033319 | DOI:10.1186/s12889-025-22040-8

BMC Musculoskelet Disord. 2025 Mar 3;26(1):217. doi: 10.1186/s12891-025-08389-2.

ABSTRACT

BACKGROUND: Osteoporosis is common in patients with knee osteoarthritis (KOA) awaiting total knee arthroplasty (TKA) and varies in different regional and ethnic. However, it is unclear whether the prevalence of osteoporosis and osteopenia in these patients is different from that in the general population. This study aims to investigate the prevalence of osteoporosis and osteopenia in both populations to help exploring the relationship between the osteoporosis and osteoarthritis, and to explore whether knee function and radiological assessments of KOA are associated with osteoporosis.

METHODS: In total, 249 patients diagnosed with KOA awaiting TKA were investigated in this cross-sectional study. The mean age was 70.9 ± 6.4 years. Bone mineral density (BMD) and T scores at the hip and lumbar spine were used to assess bone status using dual X-ray absorptiometry. A matched cohort from 2448 individuals in the Health Examination Center of our hospital was set as controls by matching sex, age (± 3.0 years) and BMI (± 1.0). The Kellgren-Lawrence grades (K-L grades), mechanical femorotibial angle (mFTA), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score and range of motion (ROM) of the knee were measured to evaluate radiological assessments and knee function in patients awaiting TKA and used to explore the association between KOA and BMD or T score. Prevalence of osteoporosis and osteopenia were investigated in the two cohorts, and inferential statistical analyses were undertaken. The chi-squared test or Fisher’s exact test was used for categorical variables while comparisons of scores were examined by ANOVA with/without Bonferroni correction or the Kruskal‒Wallis test.

RESULTS: The prevalence of osteoporosis and osteopenia in patients awaiting TKA was 30.5% (76/249) and 44.2% (110/249), respectively. In the matched cohort, 72/249 (28.9%) had osteoporosis, while 98/249 (39.4%) had osteopenia. There was no significant difference in the prevalence of osteoporosis or osteopenia between the two groups (χ2 = 2.603, P = 0.272). mFTA was significantly correlated with BMD and T score (P < 0.05), while no correlation was found between K-L grade, ROM or WOMAC and BMD or T score (P > 0.05).

CONCLUSIONS: The prevalence of osteoporosis in patients awaiting TKA was similar to that in the general population. BMD and T score were not correlated with WOMAC score or K-L grade but were correlated with mFTA.

PMID:40033308 | DOI:10.1186/s12891-025-08389-2

BMC Med Inform Decis Mak. 2025 Mar 3;25(1):109. doi: 10.1186/s12911-025-02945-5.

ABSTRACT

BACKGROUND: Algorithms and models increasingly support clinical and shared decision-making. However, they may be limited in effectiveness, accuracy, acceptance, and comprehensibility if they fail to consider patient preferences. Addressing this gap requires exploring methods to integrate patient preferences into model-based clinical decision-making.

OBJECTIVES: This scoping review aimed to identify and map applications of computational methods for incorporating patient preferences into individualized medical decision models and to report on the types of models where these methods are applied.

INCLUSION CRITERIA: This review includes articles without restriction on publication date or language, focusing on practical applications. It examines the integration of patient preferences in models for individualized clinical decision-making, drawing on diverse sources, including both white and gray literature, for comprehensive insights.

METHODS: Following the Joanna Briggs Institute (JBI) methodology, a comprehensive search was conducted across databases such as PubMed, Web of Science, ACM Digital Library, IEEE Xplore, Cochrane Library, OpenGrey, National Technical Reports Library, and the first 20 pages of Google Scholar. Keywords related to patient preferences, medical models, decision-making, and software tools guided the search strategy. Data extraction and analysis followed the JBI framework, with an explorative analysis.

RESULTS: From 7074 identified and 7023 screened articles, 45 publications on specific applications were reviewed, revealing significant heterogeneity in incorporating patient preferences into decision-making tools. Clinical applications primarily target neoplasms and circulatory diseases, using methods like Multi-Criteria Decision Analysis (MCDA) and statistical models, often combining approaches. Studies show that incorporating patient preferences can significantly impact treatment decisions, underscoring the need for shared and personalized decision-making.

CONCLUSION: This scoping review highlights a wide range of approaches for integrating patient preferences into medical decision models, underscoring a critical gap in the use of cohesive frameworks that could enhance consistency and clinician acceptance. While the flexibility of current methods supports tailored applications, the limited use of existing frameworks constrains their potential. This gap, coupled with minimal focus on clinician and patient engagement, hinders the real-world utility of these tools. Future research should prioritize co-design with clinicians, real-world testing, and impact evaluation to close this gap and improve patient-centered care.

PMID:40033306 | DOI:10.1186/s12911-025-02945-5

IEEE Trans Pattern Anal Mach Intell. 2025 Jan 9;PP. doi: 10.1109/TPAMI.2025.3527498. Online ahead of print.

ABSTRACT

Existing methods for integerized training speed up deep learning by using low-bitwidth integerized weights, activations, gradients, and optimizer buffers. However, they overlook the issue of full-precision latent weights, which consume excessive memory to accumulate gradient-based updates for optimizing the integerized weights. In this paper, we propose the first latent weight quantization schema for general integerized training, which minimizes quantization perturbation to training process via residual quantization with optimized dual quantizer. We leverage residual quantization to eliminate the correlation between latent weight and integerized weight for suppressing quantization noise. We further propose dual quantizer with optimal nonuniform codebook to avoid frozen weight and ensure statistically unbiased training trajectory as full-precision latent weight. The codebook is optimized to minimize the disturbance on weight update under importance guidance and achieved with a three-segment polyline approximation for hardware-friendly implementation. Extensive experiments show that the proposed schema allows integerized training with lowest 4-bit latent weight for various architectures including ResNets, MobileNetV2, and Transformers, and yields negligible performance loss in image classification and text generation. Furthermore, we successfully fine-tune Large Language Models with up to 13 billion parameters on one single GPU using the proposed schema.

PMID:40030978 | DOI:10.1109/TPAMI.2025.3527498

IEEE Trans Neural Syst Rehabil Eng. 2025 Jan 16;PP. doi: 10.1109/TNSRE.2025.3530154. Online ahead of print.

ABSTRACT

Assistive wearable devices can significantly enhance the quality of life for individuals with movement impairments, aid the rehabilitation process, and augment movement abilities of healthy users. However, personalizing the assistance to individual preferences and needs remains a challenge. Brain-Computer Interface (BCI) offers a promising solution for this personalization problem. The overarching goal of this study is to investigate the feasibility of utilizing passive BCI technology to personalize the assistance provided by a knee exoskeleton. Participants performed seated knee flexion-extension tasks while wearing a one-degree-of-freedom knee exoskeleton with varying levels of applied force. Their brain activities were recorded throughout the movements using electroencephalography (EEG). EEG spectral bands from several brain regions were compared between the conditions with the lowest and highest exoskeleton forces to identify statistically significant changes. A Naive Bayes classifier was trained on these spectral features to distinguish between the two conditions. Statistical analysis revealed significant increases in δ and θ activity and decreases in α and β activity in the frontal, motor, and occipital cortices. These changes suggest heightened attention, concentration, and motor engagement when the task became more difficult. The trained Naive Bayes classifier achieved an average accuracy of approximately 72% in distinguishing between the two conditions. The outcomes of our study demonstrate the potential of passive BCI in personalizing assistance provided by wearable devices. Future research should further explore integrating passive BCI into assistive wearable devices to enhance user experience.

PMID:40030934 | DOI:10.1109/TNSRE.2025.3530154

IEEE J Biomed Health Inform. 2025 Jan 15;PP. doi: 10.1109/JBHI.2025.3530107. Online ahead of print.

ABSTRACT

Employing a minimal array of electroencephalography (EEG) channels for neonatal sleep stage classification is essential for data acquisition in the Internet of Medical Things (IoMT), as single-channel and edge-based features can reduce data transfer and processing requirements, enhancing cost-effectiveness and practicality. In this paper, we evaluate the efficacy of a single channel and the viability of a binary classification scheme for discerning awake and sleep states and transitions to quiet sleep. For this, two datasets of EEG signals for neonate sleep analysis were recorded from Children’s Hospital of Fudan University, Shanghai, comprising recordings from 64 and 19 neonates, respectively. From each epoch, a diverse ensemble of 490 features was extracted through a blend of discrete and continuous wavelet transforms (DWT, CWT), spectral statistics, and temporal features. In addition, we introduced an innovative hybrid univariate and ensemble feature selection approach with multidomain feature fusion, a stacking-based ensemble classifier that outperforms existing work. We achieved 90.37%, 91.13%, and 94.88% accuracy for sleep/awake, quiet sleep/non-quiet sleep, and quiet sleep/awake, respectively. This was corroborated by significant Kappa values of 77.5%, 80.29%, and 89.76%. Using SelectPercentile, we devised three distinct feature selection mechanisms: one using DWT, one with CWT, and another incorporating both spectral and temporal features. Subsequently, SelectKBest was used to determine the most effective features. For our stacked model, we incorporated a trifecta of the ExtraTree model with variable estimators, a Random Forest, and an Artificial Neural Network (ANN) as base classifiers, and for the final prediction phase, ANN was implemented again. The model’s performance was evaluated using K-fold and leave-one-subject cross-validation.

PMID:40030895 | DOI:10.1109/JBHI.2025.3530107

IEEE Trans Pattern Anal Mach Intell. 2025 Jan 14;PP. doi: 10.1109/TPAMI.2025.3529711. Online ahead of print.

ABSTRACT

Subspace learning and Support Vector Machine (SVM) are two critical techniques in pattern recognition, playing pivotal roles in feature extraction and classification. However, how to learn the optimal subspace such that the SVM classifier can perform the best is still a challenging problem due to the difficulty in optimization, computation, and algorithm convergence. To address these problems, this paper develops a novel method named Optimal Discriminant Support Vector Machine (ODSVM), which integrates support vector classification with discriminative subspace learning in a seamless framework. As a result, the most discriminative subspace and the corresponding optimal SVM are obtained simultaneously to pursue the best classification performance. The efficient optimization framework is designed for binary and multi-class ODSVM. Moreover, a fast sequential minimization optimization (SMO) algorithm with pruning is proposed to accelerate the computation in multi-class ODSVM. Unlike other related methods, ODSVM has a strong theoretical guarantee of global convergence, highlighting its superiority and stability. Numerical experiments are conducted on thirteen datasets and the results demonstrate that ODSVM outperforms existing methods with statistical significance.

PMID:40030888 | DOI:10.1109/TPAMI.2025.3529711