Statistical power for cluster analysis

Under the surface: Low cardiac vagal tone and poor interoception in young adults with subclinical depressive symptoms

INTRODUCTION

Depressive symptoms are associated with alterations in central and autonomic nervous system activity, including misperception of bodily activity (e.g., low interoception), somatic symptoms and decreased vagally mediated heart rate variability (vmHRV). However, there is a lack of studies that examine both perception of bodily activity and autonomic function in depression. The present study investigated the association between interoception, vmHRV, and subclinical depressive symptoms.

METHOD

Eighty-eight students were enrolled and vmHRV was calculated from a 5-minute resting electrocardiogram. Interoceptive accuracy (heartbeat tracking task; heartbeat discrimination task), interoceptive sensibility (Body Perception Questionnaire), and depressive symptoms (Depression, Anxiety and Stress Scale - 21 Items) were assessed.

RESULTS

Interoceptive accuracy and sensibility positively correlated with vmHRV and negatively correlates with depressive symptoms. Cluster analysis performed on vmHRV, interoceptive accuracy, and sensibility provided two clusters: the first characterized by a pattern of low interoceptive accuracy, sensibility, and decreased resting vmHRV, the second characterized by an opposite pattern. Regression analyses showed that the first cluster was characterized by significantly higher depressive symptoms compared to the second (β = 1.97; pBonferroni = 0.04), even after controlling for sex, BMI, anxiety, and stress levels.

CONCLUSIONS

Subclinical depressive symptoms are associated with a consistent impairment in the perception and interpretation of bodily activity and altered regulatory function of the autonomic nervous system. The present results suggest that the alteration of brain-body communication could be involved in subclinical depressive symptoms. Early identification of such alterations could help with targeted preventive strategies.

Is counting a bad idea? Complex relations among children's fraction knowledge, eye movements, and performance in visual fraction comparisons

Understanding fraction magnitudes is crucial for mathematical development but is challenging for many children. Visualizations, such as tape diagrams, are thought to leverage children's early proportional reasoning skills. However, depending on children's prior knowledge, these visualizations may encourage various strategies. Children with lower fraction knowledge might rely on counting, leading to natural number bias and low performance, whereas those with higher knowledge might rely on more efficient strategies based on magnitude. This study explores the relationship between students' general fraction knowledge and their ability to visually compare fraction magnitudes represented with tape diagrams. A total of 67 children completed a fraction knowledge test and a set of comparison tasks with discretized and continuous tape diagrams while their eye movements, accuracy, and response times were recorded. Cluster analysis identified three groups. The first group, high-achieving and applying magnitude-based strategies, showed high accuracy and short response times, indicating efficiency. A second high-achieving group frequently used counting strategies, which was unexpected. This group achieved the highest accuracy but the longest response times, indicating less efficiency. The third group, low-achieving and rarely using counting strategies, had the lowest accuracy and short response times. These students tended to compare absolute sizes rather than relative sizes (i.e., showing a size bias). None of the groups exhibited a natural number bias. The study suggests that counting, although inefficient, does not necessarily lead to bias or low performance. Instead, biased reasoning with fraction visualizations can originate from reliance on absolute sizes.

No Evidence for Distinct Transcriptomic Subgroups of Devil Facial Tumor Disease (DFTD)

Contagious cancers represent one of the least understood types of infections in wildlife. Devil Facial Tumor Disease (comprised of two different contagious cancers, DFT1 and DFT2) has led to an 80% decline in the Tasmanian devil (Sarcophilus harrisii ) population at the regional level since it was first observed in 1996. There are currently no treatment options for the disease, and research efforts are focused on vaccine development. Although DFT1 is clonal, phylogenomic studies have identified different genetic variants of the pathogen. We postulated that different genetic strains may have different gene expression profiles and would therefore require different vaccine components. Here, we aimed to test this hypothesis by applying two types of unsupervised clustering (hierarchical and k-means) to 35 DFT1 transcriptomes selected from the disease's four major phylogenetic clades. The two algorithms produced conflicting results, and there was low support for either method individually. Validation metrics, such as the Gap statistic method, the Elbow method, and the Silhouette method, were ambiguous, contradictory, or indicated that our dataset only consisted of a single cluster. Collectively, our results show that the different phylogenetic clades of DFT1 all have similar gene expression profiles. Previous studies have suggested that transcriptomic differences exist between tumours from different locations. However, our study differs in that it considers both tumor purity and genotypic clade when analysing differences between DFTD biopsies. These results have important implications for therapeutic development, as they indicate that a single vaccine or treatment approach has the potential to be effective for a large cross-section of DFT1 tumors. As one of the largest studies to use transcriptomics to investigate phenotypic variation within a single contagious cancer, it also provides novel insight into this unique group of diseases.

Compressed Representation of Extreme Learning Machine with Self-Diffusion Graph Denoising Applied for Dissecting Molecular Heterogeneity

Molecular heterogeneity exists in many biological systems, such as major malignancies or diverse cell populations. Clustering of gene expression profiles has been widely used to dissect molecular heterogeneity. One drawback common to most clustering methods is that they often suffer from high dimensionality and noise, as well as feature redundancy. To address these challenges, we propose Extreme learning machine self-diffusion (ELMSD), an auto-encoder extreme learning machine feature representation method that incorporates a self-diffusion graph denoising framework to effectively dissect molecular heterogeneity. Our method, ELMSD, first learns a compressed representation of gene expression profiles from the hidden layer of the autoencoder extreme learning machine, followed by an iterative graph diffusion process to enhance the sample-to-sample similarity. The enhanced graph can largely facilitate the downstream clustering analysis, making it more efficient to analyze molecular properties. To demonstrate the utility of ELMSD, we applied it on one simulation dataset, five single-cell datasets, and 20 cancer datasets. Experiment results show that the ELMSD approach outperforms several state-of-the-art clustering methods and cancer subtypes, cell types identified by ELMSD reveal strong clinical relevance and biological interpretation. The ELMSD code is available at: https://github.com/DXCODEE/ELMSD.

Using machine learning methods to predict the outcome of psychological therapies for post-traumatic stress disorder: A systematic review

BACKGROUND

A number of treatments are available for post-traumatic stress disorder (PTSD), however, there is currently a lack of data-driven treatment selection and adaptation methods for this condition. Machine learning (ML) could potentially help to improve the prediction of treatment outcomes and enable precision mental healthcare in practice.

OBJECTIVES

To systematically review studies that applied ML methods to predict outcomes of psychological therapy for PTSD in adults (e.g., change in symptoms, dropout rate), and evaluate their methodological rigour.

METHODS

This was a pre-registered systematic review (CRD42022325021), which synthesised eligible clinical prediction studies found across four research databases. Risk of bias was assessed using the PROBAST tool. Study methods and findings were narratively synthesised, and adherence to ML best practice evaluated.

RESULTS

Seventeen studies met the inclusion criteria, including samples derived from experimental and observational study designs. All studies were assessed as having a high risk of bias, notably due to inadequately powered samples and a lack of sample size calculations. Training sample size ranged from N < 36-397. The studies applied a diverse range of ML methods such as decision trees, ensembling and boosting techniques. Five studies used unsupervised ML methods, while others used supervised ML. There was an inconsistency in the reporting of hyperparameter tuning and cross-validation methods. Only one study performed external validation.

CONCLUSIONS

ML has the potential to advance precision psychotherapy for PTSD, but to enable this, ML methods must be applied with greater adherence to best practice guidelines.

Emergence of food webs with a multi-trophic hierarchical structure driven by nonlinear trait-matching

Food webs are a central subject in community ecology, because consumption supports the flow of matter through the system, which is at the base of many of its functions. Identifying the mechanisms that are at the origin of food web structure is useful, e.g., for restoration purposes. We investigated the extent to which trait-matching, which contributes to defining the strength of trophic interactions, can cause the emergence of food webs with a non-trivial, multi-trophic, hierarchical structure. We compared for that purpose the structural properties of food webs simulated by four food web model variants, depending whether trait-matching was linear or nonlinear and whether population dynamics and evolution were accounted for (dynamical model) or not (static model). Nonlinear trait-matching can restrict interactions in phenotypic space so as to obtain localized interactions (i.e., each species interact with a small subset of species), which is a key element for food web formation. In the static case, nonlinear trait-matching allowed for the emergence of food webs, at a relatively low connectance as with random graphs. In the dynamical case, nonlinear trait-matching combined with population dynamics and evolution allowed for the formation of groups of phenotypically close species, resulting in food webs with a multi-trophic, hierarchical structure.

Motivation to exercise in patients with chronic low back pain

BACKGROUND

Chronic low back pain (CLBP) is one of the most common musculoskeletal problems worldwide. Even though regular exercise is recommended as the primary conservative approach in treating this condition, significant part of patients lead sedentary lifestyle. Motivation to exercise is one of the variables that effects the adherence of exercise-based treatments. This study aimed to characterize the motives for exercise, as posited by self-determination theory, in persons with CLBP, and to identify subgroups (clusters) of motivational profiles in combination with socioeconomic and clinical characteristics using k-means cluster analysis.

METHODS

Data were collected between September 2022 and September 2023. A total of 103 adults with CLBP completed the paper-pencil Exercise Self-Regulation Questionnaire (SRQ-E) and provided self-reported measures on anthropometric and socio-economic characteristics. Inclusion criteria were age (≥ 18 years) and non-specific CLBP (lasting longer than 12 weeks). Exclusion criteria included specific lumbar spine pathology (e.g., fracture, cancer), worsening neurological symptoms, recent injection therapy (within 3 months), and current alcohol or drug misuse.

RESULTS

Three distinct motivational clusters were identified among the 103 participants: two clusters were characterized by predominantly autonomous motivation (moderately motivated cluster: 31.1%; highly motivated cluster: 54.4%), while one cluster (controlled convinced cluster: 14.6%) showed a higher level of controlled motivation. Associations were observed between the controlled cluster and factors such as higher disability scores, longer duration of pain, greater number of completed physiotherapy sessions, and elevated BMI. Notably, the controlled motivation cluster was linked with poorer clinical outcomes.

CONCLUSIONS

This study provides insights into the exercise motivation of patients with CLBP, revealing that while most patients were primarily autonomously motivated, a notable subgroup exhibited lower, controlled motivation. The presence of controlled motivation was associated with worse functioning, longer pain duration, and increased utilization of physiotherapy services. Although these findings suggest a link between motivational profiles and clinical outcomes, the cross-sectional design limits causal inferences. Further research is needed to explore these relationships longitudinally.

TRIAL REGISTRATION

ClinicalTrials.Gov Identifier: NCT05512338 (22.8.2022, NCT05512338).

Person-centered analyses reveal that developmental adversity at moderate levels and neural threat/safety discrimination are associated with lower anxiety in early adulthood

Parsing heterogeneity in the nature of adversity exposure and neurobiological functioning may facilitate better understanding of how adversity shapes individual variation in risk for and resilience against anxiety. One putative mechanism linking adversity exposure with anxiety is disrupted threat and safety learning. Here, we applied a person-centered approach (latent profile analysis) to characterize patterns of adversity exposure at specific developmental stages and threat/safety discrimination in corticolimbic circuitry in 120 young adults. We then compared how the resultant profiles differed in anxiety symptoms. Three latent profiles emerged: (1) a group with lower lifetime adversity, higher neural activation to threat, and lower neural activation to safety; (2) a group with moderate adversity during middle childhood and adolescence, lower neural activation to threat, and higher neural activation to safety; and (3) a group with higher lifetime adversity exposure and minimal neural activation to both threat and safety. Individuals in the second profile had lower anxiety than the other profiles. These findings demonstrate how variability in within-person combinations of adversity exposure and neural threat/safety discrimination can differentially relate to anxiety, and suggest that for some individuals, moderate adversity exposure during middle childhood and adolescence could be associated with processes that foster resilience to future anxiety.

The PCOS Phenotype in Unselected Populations study: ethnic variation in population-based normative cut-offs for defining hirsutism

OBJECTIVE

Hirsutism, a diagnostic feature of polycystic ovary syndrome (PCOS), is often defined using arbitrary percentile cutoffs, rather than normative cutoffs from population-based data. We aimed to define normative cutoffs for hirsutism in diverse populations.

DESIGN

Unselected population-based cluster analysis of individual participant data (IPD).

METHODS

The PCOS Phenotype in Unselected Populations (P-PUP) study IPD asset of community-based studies, underwent k-means cluster analysis, of directly assessed hirsutism, using the modified Ferriman-Gallwey (mFG) visual scale. The primary outcome was ethnicity-specific normative cutoffs for the mFG score. Medians and cutoffs were compared across ethnic groups.

RESULTS

We included 9829 unselected, medically unbiased participants, aged 18-45 years from 12 studies conducted across 8 countries including China, Iran, Italy, Nigeria, Russia, South Korea, Turkey, and the United States. The mFG cutoff scores for hirsutism on cluster analysis varied across ethnicities, ranging from 4 to 8. White Iranians had the highest cutoff score of 8, followed by White Italians and Black Africans of 7. Asian Han Chinese, White Russian, Turkish, and Black Americans shared a cutoff of 5; White Americans, Asian Koreans, Asian Russians, and Mixed Russians shared a cutoff of 4. Comparing medians and mFG cutoffs across ethnicities confirmed the same differences.

CONCLUSION

This study confirms the 2023 International PCOS Guidelines recommendations defining hirsutism as an mFG score between 4 and 6 for the majority of populations studied, with few exceptions. However, we also highlight ethnic variation in mFG cutoff scores, suggesting that clinicians consider ethnicity in optimal diagnosis and personalized interventions.

A "Self-Milieux" perspective on help-seeking: examining the impact of a person's sociocultural background on help-seeking in people with untreated depressive symptoms

BACKGROUND

Mental illness is a global concern and the leading cause of years lived with disability. Research on help-seeking behaviour has focused on individual factors, but there is still much unexplained variance. Suggesting complex interactions between determinants of human behaviour a new framework called Self-Milieux is proposed to represent a person's sociocultural background. The article introduces a statistical approach to determine Self-Milieux and exemplarily examines its predictive validity for health-related research.

METHODS

Self-Milieux are determined through a two-stage clustering method based on the determinants socioeconomic status and self-construal profile. Descriptive analyses are used to compare Self-Milieux characteristics. Hierarchical binary logistic regression models test the association between Self-Milieux and help-seeking behaviour, while controlling for socioeconomic status as an established predictor.

RESULTS

The sample size was N = 1535 (Mage = 43.17 and 64.89% female participants). Average depression severity was M = 12.22, indicating mild to moderate symptoms. Six Self-Milieux were determined and named. Participants from privileged (aOR = 0.38) and self-sufficient (aOR = 0.37) milieux were less likely to seek help from a general practitioner than those from the entitled milieu. Participants from privileged (aOR = 0.30), collaborators (aOR = 0.50), disadvantaged (aOR = 0.33), and self-sufficient (aOR = 0.21) milieux were less likely to seek help from family members than those from the entitled and family-bound milieux.

DISCUSSION

The study's strengths and limitations, as well as the cluster methodology, are discussed. The comparative results for the six Self-Milieux are interpreted based on current research. For example, participants from some milieux follow a help-seeking process proposed in previous research, while participants from other milieux seem to show a different process, one that ends in informal help-seeking.

Assessing the environmental risks of sulfonylurea pollutants: Insights into the risk priority and structure-toxicity relationships

Sulfonylureas are widely used herbicides globally; however, the health risks associated with exposure to these compounds are poorly understood. This study used fuzzy clustering to categorize 44 sulfonylurea compounds into three risk priority levels (I, II, and III) and further investigated their structure-toxicity relationships. The order of the risk priority levels was level I＜level II＜level III. The pecking order of protein affinity was on the order of 104 M-1, which was consistent with the order of the risk priority levels. Moreover, toxic conjugations induced significant changes in protein conformation, with high-risk sulfonylurea causing substantial conformational changes. Given that the conformations of sulfonylurea within the reactive domain were highly similar, the patterns of toxic actions were considerably similar as well. Structure-toxicity relationship analysis indicated a positive correlation among Gibbs free energy change (ΔG°), affinity between sulfonylurea and protein, logarithm of the octanol-water partition coefficient (logKow), and risk priority. Specifically, a higher ΔG° value corresponded to stronger affinity, and a higher logKow value corresponded to a higher environment risk. The electronegativity of the aromatic ring on the left side of the sulfonylurea molecule is a key determinant influencing affinity - higher electronegativity of this aromatic ring weakened the affinity of sulfonylurea for protein and reduced the risk. When the aromatic ring on the left side of sulfonylurea was consistent, an increase in the electronegativity of the heterocyclic ring on the right side resulted in a stronger affinity for protein and an increased risk. This study provides a mechanistic foundation for evaluating the health risks associated with exposure to sulfonylurea.

Autoantibody clusters in childhood-onset systemic lupus erythematosus: Insights from a multicenter study with 912 patients

Objective: To identify clusters of autoantibodies in a large cSLE population and to verify possible associations between different autoantibody clusters and the following variables: demographic data, cumulative clinical and laboratory manifestations, disease activity, cumulative damage and mortality. Methods: A cross-sectional study was performed in 27 Pediatric Rheumatology University centers, including 912 cSLE patients. The frequencies of seven selected autoantibodies (anti-dsDNA, anti-Ro/SSA, anti-La/SSB, anti-Sm, anti-RNP, aCL IgM and/or IgG and LA) were used for cluster analysis using the K-means method. Results: Four distinctive antibody clusters were identified. Cluster 1 (n = 322; 35.31%) was characterized by anti-dsDNA (61.18%), low frequency of antiphospholipid antibodies (<10%), and lower frequency of cutaneous, articular manifestation (p < 0.05) and hypocomplementemia (p < 0.001) compared to the other groups. Cluster 2 (n = 158; 17.32%) comprised anti-dsDNA (93.04%), aCL (87.34%) and LA (39.87%) and higher frequencies of thrombocytopenia (p = 0.006) and antiphospholipid syndrome (p < 0.001) than the other clusters. Cluster 3 (n = 177; 19.41%) was characterized by anti-dsDNA (81.36%), anti-Sm (100%) and anti-RNP (44.63%) antibodies, as well as a higher frequency of proteinuria compared to cluster 4 (58.15% vs 56.13% vs 64.00% vs 49.80%, p = 0.031). Cluster 4 (n = 255; 27.96%) consisted of all 7 autoantibodies, with predominance of anti-dsDNA (72.55%), anti-Ro/SSA (89.8%) and anti-La/SSB (45.88%), with no specific clinical pattern, except by higher pulmonary damage (p = 0.017). Conclusions:Our study suggests that, within the context of cSLE, the coexistence of anti-dsDNA with antiphospholipid autoantibodies is linked to an elevated incidence of antiphospholipid syndrome. This association does not coincide with a proportionate increase in the occurrence of nephritis. Conversely, the cluster of anti-dsDNA with anti-Ro/SSA and anti-La/SSB antibodies was associated with pulmonary damage, requiring close surveillance.

A personalized reinforcement learning recommendation algorithm using bi-clustering techniques

Recommender systems have become a core component of various online platforms, helping users get relevant information from the abundant digital data. Traditional RSs often generate static recommendations, which may not adapt well to changing user preferences. To address this problem, we propose a novel reinforcement learning (RL) recommendation algorithm that can give personalized recommendations by adapting to changing user preferences. However, a significant drawback of RL-based recommendation systems is that they are computationally expensive. Moreover, these systems often fail to extract local patterns residing within dataset which may result in generation of low quality recommendations. The proposed work utilizes biclustering technique to create an efficient environment for RL agents, thus, reducing computation cost and enabling the generation of dynamic recommendations. Additionally, biclustering is used to find locally associated patterns in the dataset, which further improves the efficiency of the RL agent's learning process. The proposed work experiments eight state-of-the-art biclustering algorithms to identify the appropriate biclustering algorithm for the given recommendation task. This innovative integration of biclustering and reinforcement learning addresses key gaps in existing literature. Moreover, we introduced a novel strategy to predict item ratings within the RL framework. The validity of the proposed algorithm is evaluated on three datasets of movies domain, namely, ML100K, ML-latest-small and FilmTrust. These diverse datasets were chosen to ensure reliable examination across various scenarios. As per the dynamic nature of RL, some specific evaluation metrics like personalization, diversity, intra-list similarity and novelty are used to measure the diversity of recommendations. This investigation is motivated by the need for recommender systems that can dynamically adjust to changes in customer preferences. Results show that our proposed algorithm showed promising results when compared with existing state-of-the-art recommendation techniques.

Decoding Prejudice: Understanding Patterns of Adolescent Mental Health Stigma

Background/Objectives: Mental health problems are a major cause of disability, impacting nearly 20% of adolescents. Nevertheless, they are hesitant to seek help because of stigma and fear of being labelled. Adolescents often have low mental health literacy and perceive mental health problems as personal failures. To address it, our study aimed to identify subgroups within the adolescent population based on mental health knowledge, social stigma, experiences of intergroup anxiety, and endorsement of stereotypes. Methods: This cross-sectional study included 182 adolescents (50.6% male) aged 10 to 17 years (M = 13.8, SD = 2.4). Participants completed an online survey comprising the Mental Health Knowledge Schedule, Attribution Questionnaire (AQ-8-C), Intergroup Anxiety Scale, and a scale regarding stereotypes towards people with mental health problems. Cluster analysis was used to identify the subgroups. Results: We identified three subgroups: (1) "Potential Advocates", showing high mental health knowledge, low social stigma, low intergroup anxiety, and moderate endorsement of stereotypes; (2) "Ambivalents", manifesting high mental health knowledge, moderate social stigma, heightened intergroup anxiety, and low endorsement of stereotypes; and (3) "Stigmatizers", revealing low mental health knowledge, pronounced social stigma, moderate intergroup anxiety, and tendency to endorse stereotypes. Conclusions: The results highlight the multiplicity of perceptions regarding mental health and the pivotal role of knowledge, stigma, intergroup dynamics, and stereotypes in shaping attitudes. Implications for interventions targeting mental health stigma and fostering positive attitudes among adolescents are discussed, underscoring the importance of customised strategies to address the multiple needs and experiences characteristic of this developmental stage.

Neural subgroups in unaffected first-degree relatives of patients with bipolar disorder during emotion regulation

BACKGROUND

Impaired emotion regulation has been proposed as a putative endophenotype in bipolar disorder (BD). Functional magnetic resonance imaging (fMRI) studies investigating this in unaffected first-degree relatives (UR) have thus far yielded incongruent findings. Hence, the current paper examines neural subgroups among UR during emotion regulation.

METHODS

71 UR of patients with BD and 66 healthy controls (HC) underwent fMRI scanning while performing an emotion regulation task. Hierarchical cluster analysis was performed on extracted signal change during emotion down-regulation in pre-defined regions of interest (ROIs). Identified subgroups were compared on neural activation, demographic, clinical, and cognitive variables.

RESULTS

Two subgroups of UR were identified: subgroup 1 (39 UR; 55%) was characterized by hypo-activity in the dorsolateral, dorsomedial, and ventrolateral prefrontal cortex and the bilateral amygdalae, but comparable activation to HC in the other ROIs; subgroup 2 (32 UR; 45%) was characterized by hyperactivity in all ROIs. Subgroup 1 had lower success in emotion regulation compared to HC and reported more childhood trauma compared to subgroup 2 and HC. Subgroup 2 reported more anxiety, lower functioning, and greater attentional vigilance toward fearful faces compared to HC. Relatives from both subgroups were poorer in recognizing positive faces compared to HC.

CONCLUSIONS

These findings may explain the discrepancy in earlier fMRI studies on emotion regulation in UR, showing two different subgroups of UR that both exhibited aberrant neural activity during emotion regulation, but in opposite directions. Furthermore, the results suggest that impaired recognition of positive facial expressions is a broad endophenotype of BD.

Assessment of Technology Readiness in Norwegian Older Adults With Long-Term Health Conditions Receiving Home Care Services: Cross-Sectional Questionnaire Study

BACKGROUND

With the increasing number of older adults globally, there is a constant search for new ways to organize health care services. Digital health services are promising and may reduce workload and at the same time improve patient well-being. A certain level of eHealth literacy is needed to be able to use digital health services. However, knowledge of technology readiness in this target group of older adults is unclear.

OBJECTIVE

The aim of this study was to understand the technology readiness level of a group of older adults who were provided home care services in order to address the present and future needs of this group in relation to the implementation of digital health care services.

METHODS

This quantitative cross-sectional study included 149 older adults from Norway receiving home care services. The participants completed the Readiness and Enablement Index for Health Technology (READHY) instrument, assessments of well-being (World Health Organization-Five Well-Being Index [WHO-5]), and assessments of demographic and clinical variables (sex, age, education, living situation, comorbidity, use of digital devices, and use of IT). Cluster analyses were used to group the users according to their technology readiness.

RESULTS

The mean participant age was 78.6 (SD 8.0) years, and 55.7% (83/149) were women. There was good consistency within the assumed READHY scales (Cronbach α=.61-.91). The participants were grouped into 4 clusters, which differed in terms of READHY scores, demographic variables, and the use of IT in daily life. Participants in cluster 1 (n=40) had the highest scores on the READHY scales, were younger, had a larger proportion of men, had higher education, and had better access to digital devices and IT. Participants in cluster 4 (n=16) scored the lowest on eHealth literacy knowledge. Participants in cluster 1 had relatively high levels of eHealth literacy knowledge and were expected to benefit from digital health services, while participants in cluster 4 had the lowest level of eHealth literacy and would not easily be able to start using digital health services.

CONCLUSIONS

The technology readiness level varied in our cohort of Norwegian participants receiving home care. Not all elderly people have the eHealth literacy to fully benefit from digital health services. Participants in cluster 4 (n=16) had the lowest scores in the eHealth Literacy Questionnaire scales in the READHY instrument and should be offered nondigital services or would need extensive management support. The demographic differences between the 4 clusters may inform stakeholders about which older people need the most training and support to take advantage of digital health care services.

Using the alternative model of personality disorders for DSM-5 traits to identify personality types, and the relationship with disordered eating, depression, anxiety and stress

BACKGROUND

There is a substantial and growing evidence base that has identified three distinct personality types (Overcontrol, Undercontrol and Resilient) among samples of individuals with eating disorders, as well as non-clinical samples. Even in studies where up to six personality types have been identified, the three core types representing Overcontrol, Undercontrol and Resilient consistently emerge. The aim of the research was to explore whether latent Overcontrol and Undercontrol personality types could be identified using pathological personality types as part of the Alternative Model for Personality Disorders published in DSM-5. We further aimed to understand how these personality types were associated with eating pathology, depressed mood and anxiety.

METHODS

A total of 391 women, 167 men and 10 gender-diverse individuals aged 16 to 31 years completed measures of the alternative model of personality disorder traits, disordered eating behaviours, eating pathology, depression, anxiety and stress. A systematic four-step process using hierarchical, k-means, and random forest cluster analyses were used to identify the best fitting cluster solution in the data.

RESULTS

The results revealed a four-cluster solution that represented overcontrol, undercontrol, resilient and an antisocial/psychoticism cluster. The overcontrol, undercontrol, and antisocial/psychoticism types were all associated with increased disordered eating, eating pathology, depression, anxiety and stress compared to the resilient types, with the undercontrol cluster scoring significantly higher than the other three clusters on all measures of clinical pathology.

CONCLUSIONS

Pathological personality traits, as conceptualised within the DSM-5 alternative model of personality disorders may have merit for identifying overcontrol and undercontrol personality types. Our findings provide additional evidence that both overcontrol and undercontrol personality types are associated with increased eating pathology, depression, anxiety and stress.

Investigation on the Effect of Three Different Nonthermal Sterilization Methods on Volatile Organic Compounds in Safflower Using HS-GC-IMS

During the transportation, storage, and processing of safflower, it is susceptible to contamination by microorganisms, which may seriously affect the quality and safety of the flowers. Therefore, sterilization is an important step in ensuring the safety, quality, and stability of safflower products. In this study, headspace gas chromatography-ion mobility spectrometry (HS-GC-IMS) was utilized to compare the volatile organic compounds (VOCs) in safflower samples before and after sterilization with three nonthermal sterilization technologies (60Co irradiation sterilization, ultraviolet sterilization, and ozone sterilization). A total of 70 VOCs were detected in all of the safflower samples. According to the two-dimensional and three-dimensional difference contrast spectra and the fingerprint results of HS-GC-IMS, the VOCs in the safflower samples processed with the three nonthermal sterilization methods varied. By conducting principal component analysis (PCA), cluster analysis (CA), and partial least-squares regression analysis (PLS-DA) on the VOCs, it was found that 3-methyl-2-butenal, 2-heptanone, and 4-methyl-2-pentanone were the main contributors to the differences between the groups. HH-01 (not sterilized) differed significantly from HH-03 (UV sterilized) and HH-04(ozone sterilized) and differed the least from HH-02(60Co irradiation sterilized), suggesting that 60Co irradiation sterilized had the least effect on the VOCs of safflower. Therefore, 60Co irradiation technology is recommended to sterilize safflowers in large-scale production. This study provides a scientific basis for future large-scale sterilization processing of high-quality safflower. The results of this study demonstrate that HS-GC-IMS can provide strong technical support for the identification and authenticity assessment of VOCs in safflower samples.

Clinical symptoms and functional impairment in attention deficit hyperactivity disorder (ADHD) co-morbid tic disorder (TD) patients: a cluster-based investigation

BACKGROUND

Attention deficit hyperactivity disorder (ADHD) and tic disorder (TD) are two common neurodevelopmental disorders that frequently occur in childhood, and these two disorders often coexist. Cluster analysis provides a novel perspective on the heterogeneity of these commonly observed clinical disorders.

METHODS

We recruited patients with comorbid ADHD and TD from Beijing Children's Hospital between May 2022 and August 2023, collecting data on their symptoms and functional impairments. The number of clusters was determined using the elbow method, and K-means clustering was conducted. Fisher discriminant analysis and silhouette score were used for validation. Additionally, we assessed premonitory urge, strengths, and difficulties among groups. We also collected samples with ADHD alone and performed cluster analyses.

RESULTS

The number of clusters for the ADHD comorbid TD sample was determined to be two by the elbow method. The results of the cluster analysis showed that the ADHD comorbid TD sample could be divided into the severe TD group and the severe ADHD group. The severe TD group exhibits more pronounced tic symptoms, yet their age, ADHD symptoms, and functional impairment are all significantly lower than those of the severe ADHD group. Compared to samples with ADHD alone, the distribution of age and functional impairment among individuals does not change with the addition of TD symptoms, maintaining a parallel relationship with the severity of ADHD symptoms.

CONCLUSION

Patients with co-occurring ADHD and TD can be classified into two clusters based on age, symptoms, and functional impairment. In clinical interventions for these patients, while ADHD may require more attention, it is also crucial to identify the core symptoms of the patients. The heterogeneity in clinical symptom presentations highlights the need for individualized treatment approaches.

Medications and cognitive decline in Alzheimer's disease: Cohort cluster analysis of 15,428 patients

BACKGROUND

Medications for comorbid conditions may affect cognition in Alzheimer's disease (AD).

OBJECTIVE

To explore the association between common medications and cognition, measured with the Mini-Mental State Examination.

METHODS

Cohort study including persons with AD from the Swedish Registry for Cognitive/Dementia Disorders (SveDem). Medications were included if they were used by ≥5% of patients (26 individual drugs). Each follow-up was analyzed independently by performing 100 Monte-Carlo simulations of two steps each 1) k-means clustering of patients according to Mini-Mental State Examination at follow-up and its decline since previous measure, and 2) Identification of medications presenting statistically significant differences in the proportion of users in the different clusters.

RESULTS

15,428 patients (60.38% women) were studied. Four clusters were identified. Medications associated with the best cognition cluster (relative to the worse) were atorvastatin (point estimate 1.44 95% confidence interval [1.15-1.83] at first follow-up, simvastatin (1.41 [1.11-1.78] at second follow-up), warfarin (1.56 [1.22-2.01] first follow-up), zopiclone (1.35 [1.15-1.58], and metformin (2.08 [1.35-3.33] second follow-up. Oxazepam (0.60 [0.50-0.73] first follow-up), paracetamol (0.83 [0.73-0.95] first follow-up), cyanocobalamin, felodipine and furosemide were associated with the worst cluster. Cholinesterase inhibitors were associated with the best cognition clusters, whereas memantine appeared in the worse cognition clusters, consistent with its indication in moderate to severe dementia.

CONCLUSIONS

We performed unsupervised clustering to classify patients based on their current cognition and cognitive decline from previous testing. Atorvastatin, simvastatin, warfarin, metformin, and zopiclone presented a positive and statistically significant associations with cognition, while oxazepam, cyanocobalamin, felodipine, furosemide and paracetamol, were associated with the worst cluster.

Patient clusters based on demographics, clinical characteristics and cancer-related symptoms: A cross-sectional pilot study

PURPOSE

This study aimed to identify and preliminary validate distinct clusters of patients with cancer based on demographics, clinical characteristics, and symptoms and to inform future research on sample size requirements for achieving sufficient power in clustering analyses.

METHODS

This cross-sectional pilot study involved 114 patients with cancer from two hospitals in northern Italy. Data were collected on demographics, clinical characteristics, and 20 symptoms using the Edmonton Symptom Assessment System in October 2022. t-distributed stochastic neighbor embedding (t-SNE) was used to reduce the symptom data and demographics (e.g., age) into two components, which were then clustered using Ward's method. A Monte Carlo simulation was conducted based on the t-SNE components to estimate the sample size needed to achieve 80% power for different cluster solutions (k = 2, 3, 4).

RESULTS

Two distinct clusters were identified: Cluster 1 (Higher Symptom Burden Cluster) and Cluster 2 (Lower Symptom Burden Cluster). Cluster 1 patients had a higher prevalence of depression, anxiety, and drowsiness. Monte Carlo simulations indicated that 50 patients per cluster were sufficient for k = 2 clusters to achieve 80% power, whereas 90 patients per cluster were needed for k = 3 clusters and 120 patients per cluster for k = 4 clusters.

CONCLUSION

This study identified distinct patient clusters and provided preliminary evidence on the sample size required for clustering analyses in cancer research. Understanding patient clusters enables nurses to provide tailored interventions, potentially improving symptom management and overall patient care.

High-risk patient profiles for ovarian cancer: A new approach using cluster analysis of tumor markers

OBJECTIVE

Ovarian cancer remains a leading cause of cancer-related deaths in women. Early detection improves prognosis, but current diagnostic tools still need improvement. We aimed to identify high-risk patient profiles for ovarian cancer using cluster analysis of age and tumor marker data.

MATERIAL AND METHODS

A secondary dataset analysis was conducted using unsupervised learning techniques. Data were from a University Hospital, originally collected between July 2011 and July 2018 in Taiwan. In total, 349 women diagnosed with ovarian masses, including both benign and malignant tumors, were included in this analysis. The median age was 45 years, and 49 % were diagnosed with ovarian cancer in pathology. We used a hierarchical clustering algorithm to find groups of patients with similar features.

RESULTS

Two clusters were identified (N = 204 and 145), with a high-risk cluster (66.2 % malignancy) characterized by significantly older age, higher CA125, HE4, CEA, and AFP levels, and a lower CA19-9 level than the low-risk cluster (24.8 % malignancy). The assessment of clustering stability and internal validity yielded a figure of merit score of 0.970 and a silhouette coefficient of 0.524. A classification model using age, CA125, HE4, and CA19-9 demonstrated high accuracy (89.4 %), sensitivity (94.5 %), specificity (83.7 %), and a large area under the curve (89.1 %) for the risk stratification.

CONCLUSION

Integrating tumor markers with patient demographics improved the differentiation between benign and malignant ovarian masses. This approach can help clinicians prioritize high-risk patients for further diagnostic evaluation and reduce unnecessary invasive procedures for low-risk patients.

Stair-descent phenotypes in community-dwelling older adults determined using high-level balance tasks

BACKGROUND

Falls on stairs are a major cause of severe injuries among older adults, with stair descent posing significantly greater risks than ascent. Variations in stair descent phenotypes may reflect differences in physical function and biomechanical stability, and their identification may prevent falls.

AIMS

This study aims to classify stair descent phenotypes in older adults and investigate the biomechanical and physical functional differences between these phenotypes using hierarchical cluster analysis.

METHODS

Eighty-two older adults participated in this study. Stair descent was measured using a three-dimensional motion analysis system. Physical function was assessed using measures of muscle strength, walking speed, the Timed Up and Go Test (TUG), and the Community Balance and Mobility Scale (CB&M).

RESULTS

Hierarchical cluster analysis was performed on kinematic data obtained during stair descent. Three phenotypes were identified: neutral (N-type; 24%), extension (E-type; 52%), and rotation (R-type; 23%). There were no significant differences in lower limb muscle strength or walking speed among the different types, and TUG scores showed no differences in terms of mobility or balance abilities. However, CB&M scores were significantly lower for E-type and R-type compared to N-type. Sub-analyses revealed that while there were no differences in the mobility factor of CB&M between E-type and R-type, the strength factors were significantly lower compared to those for N-type.

DISCUSSION

These results suggest that E-type and R-type stair-descent patterns may be influenced by declines in standing balance ability and muscle strength.

CONCLUSIONS

These findings may inform fall-prevention training programs related to stair descent among older adults.

Machine Learning-Based Clustering Using a 12-Lead Electrocardiogram in Patients With a Implantable Cardioverter Defibrillator to Identify Future Ventricular Arrhythmia

BACKGROUND

Implantable cardioverter defibrillators (ICDs) reduce mortality associated with ventricular arrhythmia in high-risk patients with cardiovascular disease. Machine learning (ML) approaches are promising tools in arrhythmia research; however, their application in predicting ventricular arrhythmias in patients with ICDs remains unexplored. We aimed to predict and stratify ventricular arrhythmias requiring ICD therapy using 12-lead electrocardiograms (ECGs) in patients with an ICD.

METHODS AND RESULTS

This retrospective analysis included 200 adult patients who underwent ICD implantation at a single center. Patient demographics, clinical features, and 12-lead ECG data were collected. Unsupervised learning techniques, including K-means and hierarchical clustering, were used to stratify patients based on 12-lead ECG features. Dimensionality reduction methods were also used to optimize clustering accuracy. The silhouette coefficient was used to determine the optimal method and number of clusters. Of the 200 patients, 59 (29.5%) received appropriate therapy. The mean age of patients was 62.3 years, and 81.0% were male. The mean follow-up period was 2,953 days, with no significant intergroup differences. Hierarchical clustering into 3 clusters proved to be the most accurate (silhouette coefficient=0.585). Kaplan-Meier curves for these 3 clusters revealed significant differences (P=0.026).

CONCLUSIONS

We highlight the potential of ML-based clustering using 12-lead ECGs to help in the risk stratification of ventricular arrhythmia. Future research in a larger multicenter setting may provide further insights and refine ICD indications.

Profiling Autism and Attention Deficit Hyperactivity Disorder Traits in Children with SYNGAP1-Related Intellectual Disability

SYNGAP1-related ID is a genetic condition characterised by global developmental delay and epilepsy. Individuals with SYNGAP1-related ID also commonly show differences in attention and social communication/interaction and frequently receive additional diagnoses of Autism Spectrum Disorder (ASD) or Attention Deficit Hyperactivity Disorder (ADHD). We thus set out to quantify ASD and ADHD symptoms in children with this syndrome. To assess ASD and ADHD, parents and caregivers of a child with SYNGAP1-related ID (N = 34) or a typically developing control (N = 21) completed the Social Responsiveness Scale-2, the Social Communication Questionnaire with a subset of these also completing the Conners-3. We found that those with SYNGAP1-related ID demonstrated higher levels of autistic traits on both the SRS and SCQ than typically developing controls. On the SRS, those with SYNGAP1-related ID scored highest for restricted repetitive behaviours, and were least impaired in social awareness. On the Conners-3, those with SYNGAP1-related ID also showed a high prevalence of ADHD traits, with scores demonstrating difficulties with peer relations but relatively low occurrence of symptoms for DSM-5 conduct disorder and DSM-5 oppositional defiant disorder. Hierarchical clustering analysis highlighted distinct SYNGAP1-related ID subgroups for both ASD and ADHD traits. These findings provide further characterisation of the SYNGAP1-related ID behavioural phenotype, guiding diagnosis, assessment and potential interventions.

Machine learning identifies clusters of the normal adolescent spine based on sagittal balance

PURPOSE

This study applied a machine learning semi-supervised clustering approach to radiographs of adolescent sagittal spines from a single pediatric institution to identify patterns of sagittal alignment in the normal adolescent spine. We sought to explore the inherent variability found in adolescent sagittal alignment using machine learning to remove bias and determine whether clusters of sagittal alignment exist.

METHODS

Multiple semi-supervised machine learning clustering algorithms were applied to 111 normal adolescent sagittal spines. Sagittal parameters for resultant clusters were determined.

RESULTS

Machine learning analysis found that the spines did cluster into distinct groups with an optimal number of clusters ranging from 3 to 5. We performed an analysis on both 3 and 5-cluster groups. The 3-cluster groups analysis found good consistency between methods with 96 of 111, while the analysis of 5-cluster groups found consistency with 105 of 111 spines. When assessing for differences in sagittal parameters between the groups for both analyses, there were differences in T4-12 TK, L1-S1 LL, SS, SVA, PI-LL mismatch, and TPA. However, the only parameter that was statistically different for all groups was SVA.

CONCLUSIONS

Based on machine learning, the adolescent sagittal spine alignments do cluster into distinct groups. While there were distinguishing features with TK and LL, the most important parameter distinguishing these groups was SVA. Further studies may help to understand these findings in relation to spinal deformities.

Features influencing surface acting of different clusters of nursing students in vocational college based on interpretable machine learning: A cross-sectional study

AIM

To explore and explain the mechanisms that influence surface acting in nursing students with different characteristics.

BACKGROUND

Nurses are now expected to deliver patient-centered care which necessitates the emotional labor. Surface acting, a form of emotional labor, can lead to negative outcomes. Given that nursing students are the backbone of the future nursing profession, there is an urgent need to investigate their surface acting tendencies and identify potential factors for early intervention.

DESIGN

A cross-sectional study.

METHODS

This study was surveyed in a vocational college in Gansu, China. Participants completed the general information questionnaire, Bem Sex Role Inventory, Professional Identity Questionnaire of Nursing Students and Surface Acting Scale. K-means cluster analysis was performed, followed by random forest algorithm and SHapley Additive exPlanations based on Python program.

RESULTS

A total of 1241 nursing students from vocational college were investigated and were clustered into 4 groups. The five dimensions of professional identity had higher feature importance in all four groups, with professional self-image having the highest feature importance in Cluster 3. Professional self-image and understanding retention benefits and turnover risks were negative predictors of surface acting in all four groups. Social comparison and self-reflection, independence of career choice and social modeling regarding nursing profession were positively correlated with surface acting in specific groups. In Cluster 1, there exists a positive correlation between professional self-image and the constructs of social comparison and self-reflection; as well as a negative correlation between maternal education and understanding of retention benefits and turnover risks.

CONCLUSIONS

Professional identity significantly influences surface acting behaviors among nursing students, with professional self-image serving as a key negative predictor. Positive family conditions, access to educational resources, parental literacy, masculine or feminine gender roles and first-year nursing students, these traits have implications when dimensions of professional identity are used to predict surface acting behaviors.

Distinct patient characteristics are associated with clinical presentation and prognosis in thromboangiitis obliterans

Background: Thromboangiitis obliterans (TAO) is a rare but threatening disease associated with significant morbidity and mortality. The pathophysiology is poorly understood, the diagnosis is often obscure and causal treatment options are limited. In the current study, we aimed to identify distinct TAO patient clusters that differed in clinical presentation and prognosis. Patients and methods: We retrospectively analysed a cohort of 48 patients with the working diagnosis TAO who were assessed for clinical presentation at hospital admission. We applied hierarchical clustering to divide patients into clinically meaningful subgroups. Results: Patients were followed-up for a median of 95 months. We found that cluster analyses including a variety of demographic and diagnostic parameters were valuable to identify patient subgroups with similar clinical presentation, but with different clinical course of the disease, including the individual risk for mortality and major amputation. Patients treated with statins showed a significantly better survival, which may allow us to hypothesize that a conventional secondary prevention strategy, which is recommended for atherosclerotic artery diseases, may be of benefit also in patients that present with TAO. Conclusions: The current data may help to develop strategies to identify high-risk TAO patients. Furthermore, statins may serve as a readily available therapeutic option to this rare but serious disease.

Differential Treatment Responses of Maltreated and Neglected Children and Adolescents Following an Evidence-based Multisystemic Intervention

Limited studies have investigated differential treatment responses to family-based treatment programs and subgroup trajectories in youth in a high-risk context. This study pioneered an examination of Multisystemic Therapy for Child Abuse and Neglect (MST-CAN) and built on prior research that identified subgroups with different psychopathologies. Participants included 208 parent-child dyads enrolled in the MST-CAN evaluation in Switzerland. Parents reported their children's (Mage = 10.27 years, SDage = 3.5, 44.2% girls, 55.8% boys, 98.6% White) emotional and behavioral problems. Longitudinal data were examined to analyze the differential changes within the pre- and post-treatment (T1 and T2) subgroups. The T1 cluster and T2 cluster were cross-tabulated to examine changes in the symptom class over time. Overall, the treatment proved to be highly beneficial. Subgroup analyses revealed that four out of the five subgroups (80%) showed positive changes in at least two outcome measures. The treatment was most successful for children with externalizing symptoms. Children with multiple symptoms also showed improvements across different symptoms. Regarding specific symptoms, children with anxious-avoidant symptoms benefited from the treatment. Additionally, the treatment was beneficial for children with normative emotions and behavior. Meanwhile, the treatment did not have any significant effects for children with internalizing symptoms. Notably, child neglect was reduced in three (60%) subgroups. The symptom class remained stable across time for children with externalizing and multiple symptoms. Ultimately, MST-CAN reduced emotional and behavioral problems and child neglect in most families. Understanding children's differential treatment responses to complex treatment programs is essential to adequately address different needs.

Heavy metals in sediments of the river-lake system in the Dianchi basin, China: Their pollution, sources, and risks

The accumulation of heavy metals in river and lake sediments in basins seriously threatens ecological safety and human health. To manage the basin effectively, it is crucial to understand pollution levels and identify and quantify the sources and risks of heavy metals in rivers and lakes separately for targeted control. In this study, 34 sediment samples were collected from the Dianchi Basin, China, and the pollution, sources, and risks in the river-lake system sediments were systematically analysed for cadmium (Cd), chromium (Cr), arsenic (As), mercury (Hg), lead (Pb), copper (Cu), zinc (Zn), and nickel (Ni). The results showed that at least half of the heavy metals in the lakes and rivers exceeded the local soil background values during the flood and dry seasons. Heavy metal concentrations were generally higher in the lakes, with high concentrations in the lakes and nearby rivers. Through positive matrix factorisation and Geodetector traceability discovery, agricultural activities were found to be the main source of heavy metals in river sediments, whereas urban activities were the main source in lake sediments. Ecological risk assessments indicated that Hg and Cd were the main risk factors, causing pollution in lakes due to atmospheric deposition and traffic emissions and moderate pollution in rivers due to atmospheric deposition and agricultural production. Health risk assessments revealed that As and Ni were the main carcinogenic risks, originating from human and industrial activities in lakes, and from agricultural and natural sources in rivers. Children faced higher carcinogenic risks than adults, possibly because of their behaviour and physiology. Overall, the presence of heavy metals, along with their ecological and health risks, was significantly higher in the lakes than in the rivers. This study provides a comprehensive overview of the pollution, sources, and risks of eight heavy metals in the river-lake system of the Dianchi Basin.

Detecting cardiovascular diseases using unsupervised machine learning clustering based on electronic medical records

BACKGROUND

Electronic medical records (EMR)-trained machine learning models have the potential in CVD risk prediction by integrating a range of medical data from patients, facilitate timely diagnosis and classification of CVDs. We tested the hypothesis that unsupervised ML approach utilizing EMR could be used to develop a new model for detecting prevalent CVD in clinical settings.

METHODS

We included 155,894 patients (aged ≥ 18 years) discharged between January 2014 and July 2022, from Xuhui Hospital, Shanghai, China, including 64,916 CVD cases and 90,979 non-CVD cases. K-means clustering was used to generate the clustering models with k = 2, 4, and 8 as predetermined number of clusters k = 2, 4, and 8. Bayesian theorem was used to estimate the models' predictive accuracy.

RESULTS

The overall predictive accuracy of the 2-, 4-, and 8-classification clustering models in the training set was 0.856, 0.8634, and 0.8506, respectively. Similarly, the predictive accuracy of the 2-, 4-, and 8-classification clustering models in the testing set was 0.8598, 0.8659, and 0.8525, respectively. After reducing from 19 dimensions to 2 dimensions by principal component analysis, significant separation was observed for CVD cases and non-CVD cases in both training and testing sets.

CONCLUSION

Our findings indicate that the utilization of EMR data can support the development of a robust model for CVD detection through an unsupervised ML approach. Further investigation using longitudinal design is needed to refine the model for its applications in clinical settings.

Diversity analysis, nutrition, and flavor evaluation of amino acids in Chinese native geese germplasms

Background and Aim

As living standards improve and consumption patterns shift, the market for goose meat continues to grow because of its exceptional dietary quality and distinctive flavor. The composition and content of amino acids are critical for determining the nutritional value and flavor of meat. This study aimed to evaluate the nutritional value and flavor of 10 Chinese native geese germplasms based on their amino acid content and composition.

Materials and Methods

A total of 568 geese from 10 Chinese native geese germplasms reared under identical conditions were slaughtered at 10 weeks of age. The pectoralis and thigh muscles (thighs) were collected to determine the amino acid content using an amino acid analyzer. Subsequently, diversity, variance, cluster, and principal component analyses were performed to identify superior germplasm with improved nutrition and flavor.

Results

The results revealed 17 amino acids in goose meat, with Glutamate and Aspartate being the most abundant. The amino acid scores of goose meat exceeded the values recommended by the Food and Agriculture Organization/World Health Organization. The Shannon-Wiener Diversity Index (1.72-2.07) indicated a high degree of diversity in amino acid content among geese germplasms. The pectoralis exhibited significantly higher amino acid content (p < 0.05 or p < 0.01) than the thigh, except for the essential amino acids to total amino acids ratio (p < 0.05 or p < 0.01). The 10 germplasms were categorized into four clusters, with Wanxi (WX) and Taizhou (TZ) geese grouped in Cluster I, displaying significantly higher nutritional value and flavor (p < 0.05 or p < 0.01) than other germplasms.

Conclusion

Germplasms with superior nutritional value and flavor (WX and TZ) were identified among 10 Chinese native geese germplasms, providing valuable insights for the conservation of existing germplasms and the cultivation of new goose breeds with improved meat quality.

Early trajectories and moderators of autistic language profiles: A longitudinal study in preschoolers

LAY ABSTRACT

Language development can greatly vary among autistic children. Children who struggle with language acquisition often face many challenges and experience lower quality of life. However, little is known about the early language trajectories of autistic preschoolers and their moderators. Autistic language can be stratified into three profiles. Language unimpaired experience little to no language difficulties; language impaired show significant difficulties in language; minimally verbal never develop functional language. In this study, we used a longitudinal sample of preschoolers with autism and with typical development (aged 1.5-5.7 years). We replicated the three language profiles through a data-driven approach. We also found that different factors modulated the language outcome within each group. For instance, non-verbal cognition at age 2.4 moderated the participants' attribution to each language profile. Moreover, early intervention moderated verbal outcome in the language impaired profile. In conclusion, we provided a detailed description of how autistic preschoolers acquire language, and what factors might influence their trajectories. Our findings could inspire more personalized intervention for early autistic language difficulties.

A Lifeline to Fill the Silence of Homelessness: Person-Centered Analysis of Digital Coping and Links to Mental and Physical Health

Individuals experiencing homelessness are among the most vulnerable population for mental and physical health disparities. Despite navigating numerous stressors on a day-to-day basis, they are vastly underrepresented within coping research. Using a person-centered approach, this study addresses ways in which technology is leveraged to manage ongoing stressors associated with the experience of homelessness. We employed a two-step and k-means cluster analysis within a sample of unhoused individuals (n = 66). Two distinct clusters emerged, revealing unique patterning of digital coping, stress, self-efficacy, and technology use. Resulting clusters were validated across numerous health outcomes, including mental and physical health problems, as well as digital service use and experience of homelessness. High digital coping/low self-efficacy individuals (65% of sample) reported high levels of digital self-efficacy, yet lower levels of general self-efficacy. In contrast, low digital engagement/high self-efficacy individuals (35% of sample) engaged in relatively lower digital coping and technology use, with lower stress and higher general self-efficacy. High digital coping/low self-efficacy individuals, in turn, reported more mental and physical health problems; whereas low digital engagement/high self-efficacy reported somewhat decreased digital access. Relatively few differences emerged between the clusters on experiences of homelessness. Due to the transient nature of unhoused people, reaching such vulnerable populations via technology to support their digital coping and subsequently enhance well-being outcomes represents a critical next step for digital equity. This population is poised to benefit from digital equity efforts, with critical implications for reduced health disparities.

Clustering Methods in Rheumatic and Musculoskeletal Disease Research: An Educational Guide to Best Research Practices

Clinical manifestations and disease progression often exhibit significant variability among patients with rheumatic diseases, complicating diagnosis and treatment strategies. A better understanding of disease heterogeneity may allow for personalized treatment strategies. Cluster analysis is a class of statistical methods that aims to identify subgroups or patterns within a dataset. Cluster analysis is a type of unsupervised learning, meaning there are no outcomes or labels to guide the analysis (ie, there is no ground truth). This makes it difficult to assess the accuracy or validity of the identified clusters, and these methods therefore require thoughtful planning and careful interpretation. Here, we provide a high-level overview of clustering, including different types of clustering methods and important considerations when undertaking clustering, and review some examples from the rheumatology literature.

The discrepancies in parent and teacher reports of children's behavioral inhibition provide domain-specific information about psychopathology and parenting

Introduction

Behavioral inhibition is a temperamental factor that increases the risk of internalizing disorders. Therefore, the identification of highly inhibited children is of great importance. However, informant discrepancies make this process difficult. In a cluster analytic approach, we aimed to use both parent and teacher reports of behavioral inhibition in order to gain a more detailed picture about children's behavioral inhibition in different contexts and to characterize highly inhibited children.

Methods

Parents and teachers of 318 preschool children completed a questionnaire, which included the Behavioral Inhibition Questionnaire (BIQ) and the Strengths and Difficulties Questionnaire (SDQ). Parents also reported their parenting behavior on the Multidimensional Assessment of Parenting Questionnaire (MAPS). A two-step cluster analysis was conducted on BIQ parent and teacher reports, and the resulting clusters were compared on the SDQ externalizing and internalizing subscales. Multinomial logistic regression analyses were conducted separately for girls and boys to predict cluster membership based on the MAPS hostility, lax control and physical control subscales.

Results

Four clusters were identified, labelled as medium-low (ML), low-elevated (LE), elevated-elevated (EE) and high-high (HH), based on the levels of BIQ parent and teacher reports, respectively. In the HH cluster, mean scores of the SDQ internalizing subscales as reported by parents and teachers were significantly higher, and in boys but not in girls, mean scores of the SDQ externalizing subscale as reported by teachers were lower than in the other clusters. High levels of hostility predicted group membership of HH compared to LE and EE in both genders. Furthermore, in boys, lax control and physical control were also found to be significant when comparing HH to EE and LE, respectively.

Discussion

Our results suggest that the joint use of parent and teacher reports on behavioral inhibition may increase the ability to identify highly inhibited children at risk of developing internalizing disorders and add to our understanding of the underpinnings of children's inhibited behavior in different contexts.

Unsupervised Clustering in Neurocritical Care: A Systematic Review

Managing patients with acute brain injury in the neurocritical care (NCC) unit has become increasingly complex because of technological advances and increasing information derived from multiple data sources. Diverse data streams necessitate innovative approaches for clinicians to understand interactions between recorded variables. Unsupervised clustering integrates different data streams and could be supportive. Here, we provide a systematic review on the use of unsupervised clustering using NCC data. The primary objective was to provide an overview of clustering applications in NCC studies. As a secondary objective, we discuss considerations for future NCC studies. Databases (Medline, Scopus, Web of Science) were searched for unsupervised clustering in acute brain injury studies including traumatic brain injury (TBI), subarachnoid hemorrhage, intracerebral hemorrhage, acute ischemic stroke, and hypoxic-ischemic brain injury published until  March 13th 2024. We performed the systematic review in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines. We identified 18 studies that used unsupervised clustering in NCC. Predominantly, studies focused on patients with TBI (12 of 18 studies). Multiple research questions used a variety of resource data, including demographics,  clinical- and monitoring data, of which intracranial pressure was most often included (8 of 18 studies). Studies also covered various clustering methods, both traditional methods (e.g., k-means) and advanced methods, which are able to retain the temporal aspect. Finally, unsupervised clustering identified novel phenotypes for clinical outcomes in 9 of 12 studies. Unsupervised clustering can be used to phenotype NCC patients, especially patients with TBI, in diverse disease stages and identify clusters that may be used for prognostication. Despite the need for validation studies, this methodology could help to improve outcome prediction models, diagnostics, and understanding of pathophysiology.Registration number: PROSPERO: CRD4202347097676.

Analysis of the Prevalence of Bacterial Pathogens and Antimicrobial Resistance Patterns of Edwardsiella piscicida in Largemouth Bass (Micropterus salmoides) from Guangdong, China

To gain insights into the prevalence and antimicrobial resistance patterns of major bacterial pathogens affecting largemouth bass (Micropterus salmoides) in the Pearl River Delta (PRD) region, Guangdong, China, a study was conducted from August 2021 to July 2022. During this period, bacteria were isolated and identified from the internal organs of diseased largemouth bass within the PRD region. The antimicrobial resistance patterns of 11 antibiotics approved for use in aquaculture in China were analyzed in 80 strains of Edwardsiella piscicida using the microbroth dilution method. The results showed that 151 bacterial isolates were obtained from 532 samples, with E. piscicida (17.29%, 92/532), Aeromonas veronii (4.70%, 25/532), and Nocardia seriolae (2.26%, 12/532) being the main pathogens. Notably, E. piscicida accounted for the highest proportion of all isolated bacteria, reaching 60.92% (92/151), and mainly occurred from November to April, accounting for 68.48% (63/92) of the cases. The symptoms in largemouth bass infected with E. piscicida included ascites, enteritis, and hemorrhaging of tissues and organs. The drug sensitivity results showed that the resistance rates of all E. piscicida strains to ciprofloxacin, all sulfonamides, thiamphenicol, florfenicol, enrofloxacin, doxycycline, flumequine, and neomycin were 96.25%, 60-63%, 56.25%, 43.75%, 40%, 32.5%, 16.25%, and 1.25%, respectively. In addition, 76.25% (61/80) of these strains demonstrated resistance to more than two types of antibiotics. Cluster analysis revealed 23 antibiotic types (A-W) among the 80 isolates, which were clustered into two groups. Therefore, tailored antibiotic treatment based on regional antimicrobial resistance patterns is essential for effective disease management. The findings indicate that in the event of an Edwardsiella infection in largemouth bass, neomycin, doxycycline, and flumequine are viable treatment options. Alternatively, one may choose drugs that are effective as determined by clinical drug sensitivity testing.

Principal component analysis biplot visualization of electromyogram features for submaximal muscle strength grading

BACKGROUND

Submaximal muscle strength grading is clinically significant to monitor the progress of rehabilitation. Especially muscle strength grading of core back muscles is challenging using the conventional manual muscle testing (MMT) methods. The muscles are crucial to recovery from back pain, spinal cord injury, stroke and other related diseases. The subjective nature of MMT, adds more ambiguity to grade fine progressions in submaximal strength levels involving 4-, 4 and 4+ grades. Electromyogram (EMG) has been widely used as a quantitative measure to provide insight into the progress of muscle strength. However, several EMG features have been reported in previous studies, and the selection of suitable features pertaining to the problem has remained a challenge.

METHOD

Principal Component Analysis (PCA) biplot visualization is employed in this study to select EMG features that highlight fine changes in muscle strength spanning the submaximal range. Features that offer maximum loading in the principal component subspace, as observed in the PCA biplot, are selected for grading submaximal strength. The performance of the proposed feature set is compared with conventional Principal Component (PC) scores. Submaximal muscle strength grades of 4-, 4, 4+ or 5 are assigned using K-means and Gaussian mixture model clustering methods. Clustering performance of the two feature selection methods is compared using the silhouette score metric.

RESULTS

The proposed feature set from biplot visualization involving Root Mean Square (RMS) EMG and Waveform Length in combination with Gaussian Mixture Model (GMM) clustering method was observed to offer maximum accuracy. Muscle-wise mean Silhouette Index (SI) scores (p < 0.05) of .81, .74 (Longissimus thoracis left, right) and .73, .77 (Iliocostalis lumborum left, right) were observed. Similarly grade wise mean SI scores (p < 0.05) of .80, .76, .73, and .981 for grades 4-, 4, 4+, and 5 respectively, were observed.

CONCLUSION

The study addresses the problem of selecting minimum features that offer maximum variability for EMG assisted submaximal muscle strength grading. The proposed method emphasizes using biplot visualization to overcome the difficulty in choosing appropriate EMG features of the core back muscles that significantly distinguishes between grades 4-, 4, 4+ and 5.

Health-related quality of life of people with type 1 diabetes: An IMI2 SOPHIA post hoc analysis of FUTURE and ADJUNCT-ONE

AIM

To characterize and stratify health-related quality of life in individuals with type 1 diabetes (T1D) using body mass index (BMI) and clustering analysis.

MATERIAL AND METHODS

Baseline data on individuals with T1D were pooled from two studies. A post hoc analysis of health-related quality of life, measured using the 36-item Short-Form questionnaire, was performed, referenced to the 2010 US general population. Descriptive statistics were presented for the pooled cohort and per BMI category. K-means clustering was performed. One-way analysis of variance was conducted to examine differences in clinical characteristics between clusters.

RESULTS

The pooled cohort consisted of 2256 individuals with T1D (age: 45.4 ± 15.0 years, BMI: 26.2 ± 4.6 kg/m2, diabetes duration: 22.7 ± 13.5 years). All quality-of-life domains were slightly lower than 50(the general population's mean), except for vitality. Individuals with a BMI ≥30 kg/m2 reported lower scores for bodily pain, physical functioning, general health, and vitality. A first cluster with a high and a second cluster with a low quality of life were identified, with significant differences in the mental (Cluster 1: 53.8 ± 6.8 vs. Cluster 2: 39.5 ± 10.7; p < 0.001) and physical component summary scores (Cluster 1: 49.6 ± 6.3 vs. Cluster 2: 35.2 ± 12.0; p < 0.001), which exceeded differences found between BMI categories.

CONCLUSIONS

In our population of people living with T1D, higher BMI may have adversely impacted physical domains of quality of life, but larger differences between the high- and low-quality-of-life cluster indicate that more factors play a role.

Cluster analysis in fibromyalgia: a systematic review

BACKGROUND

The multifaceted nature of Fibromyalgia syndrome (FM) symptoms has been explored through clusters analysis.

OBJECTIVE

To synthesize the cluster research on FM (variables, methods, patient subgroups, and evaluation metrics).

METHODS

We performed a systematic review following the PRISMA recommendations. Independent searches were performed on PubMed, Embase, Web of Science, and Cochrane Central, employing the terms "fibromyalgia" and "cluster analysis". We included studies dated to January 2024, using the cluster analysis to assess any physical, psychological, clinical, or biomedical variables in FM subjects, and descriptively synthesized the studies in terms of design, cluster method, and resulting patient profiles.

RESULTS

We included 39 studies. Most with a cross-sectional design aiming to classify subsets based on the severity, adjustment, symptomatic manifestations, psychological profiles, and response to treatment, based on demographic and clinical variables. Two to four different profiles were found according to the levels of severity and adjustment to FMS. According to symptom manifestation, two to three clusters described the predominance of pain versus fatigue, and thermal pain sensitivity (less versus more sensitive). Other clusters revealed profiles of personality (pathological versus non-pathological) and psychological vulnerability (suicidal ideation). Additionally, studies identified different responses to treatment (pharmacological and multimodal).

CONCLUSION

Several profiles exist within FMS population, which point out to the need for specific treatment options given the different profiles and an efficient allocation of healthcare resources. We notice a need towards more objective measures, and the validation of the cluster results. Further research might investigate some of the assumptions of these findings, which are further discussed in this paper.

A cluster analysis of attachment styles in patients with borderline personality disorder, bipolar disorder and ADHD

BACKGROUND

Insecure adult attachment has been associated with psychiatric disorders characterized by emotional dysregulation, such as borderline personality disorder (BPD), bipolar disorder (BD) and attention deficit/hyperactivity disorder (ADHD). However, little is known about the differences in attachment patterns between these diagnostic groups. The aim of this study was to identify clusters of adult attachment style in a cross-diagnostic group of patients with BDP and/or BD and/or ADHD and explore the characteristics of these clusters based on temperament profile, childhood trauma and psychiatric diagnoses.

METHODS

K-means cluster analysis was used to identify subgroups, based on the Attachment Style Questionnaire Short Form dimensions, in a clinical cohort of 150 young adults (113 women and 37 men, mean age ± SD = 23.3 ± 2.1) diagnosed with BPD, and/or BD, and/or ADHD.

RESULTS

Three distinct clusters were identified: a secure, an insecure/avoidant-anxious and an insecure/avoidant cluster. These three clusters differed in temperament profile and related psychiatric diagnoses.

CONCLUSIONS

The three clusters of attachment in individuals with BPD, BD and/or ADHD could support differentiation between the disorders as well provide information usable for planning of treatment.

Characteristics and predictors of persistent somatic symptoms in patients with cardiac disease

Persistent somatic symptoms (PSS) are a diagnostic core criterion of the somatic symptom disorder. This longitudinal study aims to determine the frequency of PSS in patients with cardiac disease, identify potential predictive factors, and investigate its impact on healthcare utilization. Somatic symptoms were assessed with the Somatic Symptom Scale-8 four times over the course of three months in consecutively approached cardiac outpatients. Patients were grouped having PSS vs. not having PSS following a psychometric-driven approach based on the SSS-8 cut-off score and a data-driven approach applying cluster analysis. T-tests were performed to compare the characteristics between patients having vs. not having PSS. To identify predictors of group affiliation, we conducted multivariable logistic regressions. Additionally, analyses of covariance were used to further examine associations between healthcare utilization and group affiliation. The study included 95 patients (30.5% female) with a mean age of 60.5 years (SD = 8.7). All patients had at least one of the following cardiac diseases recorded in their medical history: coronary heart disease (n = 51), myocardial infarction (n = 21), valve disease (n = 22), cardiomyopathy (n = 15), cardiac dysrhythmia (n = 43), and heart failure (n = 12). 30 (32%) were grouped having PSS according to the psychometric-driven approach and 27 (28%) according to the data-driven approach. For both approaches, patients with PSS were more likely to be female, unemployed, reporting angina pectoris, having higher depression, and higher anxiety severity (for all: p ≤ 0.05). Predictors of PSS group affiliation were female gender, higher age, depression severity, and angina pectoris (for all: p ≤ 0.015). Patients with PSS more frequently visited general practitioners and cardiologists compared to patients without PSS (p ≤ 0.013). Enhancing our knowledge of PSS in patients with cardiac disease could help to improve identification of patients' specific needs and the factors to consider in diagnosis and individualized treatment.

Using machine learning methods to identify trajectories of change and predict responders and non-responders to short-term dynamic therapy

OBJECTIVES

Predicting therapy responders can significantly improve clinical outcomes. This study aims to identify predictors of response to short-term dynamic therapy.

METHODS

Data from 95 patients who underwent 16-session therapy were analyzed using machine learning. Weekly progress was monitored with the Outcome Questionnaire (OQ45) and Target Complaints (TC). A machine learning model identified change trajectories for responders and non-responders, with a random forest algorithm and elastic net modeling predicting trajectory group membership using pre-treatment data.

RESULTS

A weak positive relationship was found between the trajectories of the two outcome variables. The results of the different analysis methods were compared and discussed. Important predictors of OQ45 trajectories, based on random forest modeling, included initial symptom severity, difficulties in emotion regulation, coldness, avoidant attachment, conscientiousness, interpersonal problems, non-acceptance of negative emotion, neuroticism, emotional clarity, impulsivity, and emotion awareness (72.8% accuracy). Initial problem severity, self-scarifying extraversion, and non-assertiveness were the most dominant predictors for TC trajectories (62.8% accuracy).

CONCLUSIONS

These findings offer data-driven insights for selecting short-term dynamic therapy. Predicting response for the OQ45, a nomothetic measure, does not extend to the TC, an idiographic measure, and vice versa, highlighting the importance of multidimensional outcome evaluations for personalized treatment.

Identifying distinct phenotypes of patients with juvenile systemic lupus erythematosus: results from a cluster analysis by the Egyptian college of rheumatology (ECR) study group

PURPOSE

Juvenile systemic lupus erythematosus (J-SLE) is a complex, heterogeneous disease affecting multiple organs. However, the classification of its subgroups is still debated. Therefore, we investigated the aggregated clinical features in patients with J-SLE using cluster analysis.

METHODS

Patients (≤ 16 years) diagnosed using the Systemic Lupus International Collaborating Clinics (SLICC) classification criteria were identified from the clinical database of the Egyptian College of Rheumatology (ECR) SLE study group. Demographic data, clinical characteristics, laboratory features, and current therapies were selected. A cluster analysis was performed to identify different clinical phenotypes.

RESULTS

Overall, 404 patients, of whom 355 (87.9%) were female, had a mean age at diagnosis of 11.2 years and a mean disease duration of 2.3 years. We identified four distinct subsets of patients. Patients in cluster 1 (n = 103, 25.5%) were characterized predominantly by mucocutaneous and neurologic manifestations. Patients in cluster 2 (n = 101, 25%) were more likely to have arthritis and pulmonary manifestations. Cluster 3 (n = 71, 17.6%) had the lowest prevalence of arthritis and lupus nephritis (LN), indicative of mild disease intensity. Patients in cluster 4 (n = 129, 31.9%) have the highest frequency of arthritis, vasculitis, and LN. Cluster 1 and 4 patients had the highest disease activity index score and were less likely to use low-dose aspirin (LDA). The SLE damage index was comparable across clusters.

CONCLUSIONS

Four identified J-SLE clusters express distinct clinical phenotypes. Attention should be paid to including LDA in the therapeutic regimen for J-SLE. Further work is needed to replicate and clarify the phenotype patterns in J-SLE.

Decline in and recovery of fertility rates after COVID-19-related state of emergency in Japan

INTRODUCTION

The COVID-19 pandemic led to a decline in fertility rates worldwide. Although many regions have experienced a temporary drop in fertility rates with the spread of the infection, subsequent recovery has varied across countries. This study aimed to evaluate the impact of COVID-19 infection rates and regional sociodemographic factors on the recovery of fertility rates in Japan following the state of emergency.

METHODS

This study examined prefectural fertility data from before the COVID-19 pandemic to forecast fertility rates up to 2022 using a seasonal autoregressive integrated moving average model. A regression analysis was conducted on fertility rates during the first state of emergency and the subsequent recovery rate with respect to the number of new COVID-19 cases and sociodemographic factors specific to each prefecture.

RESULTS

During the first state of emergency, the monthly fertility rate decreased by an average of -13.8% (SD: 6.26, min: -28.78, max: 0.15) compared with the previous year. Over the following 22 months, the average fertility recovery rate was +2.31% (SD: 3.57; min: -8.55, max: 19.54). Multivariate analysis of the impact of the pandemic on fertility changes during the first emergency indicated a negative correlation between new COVID-19 cases per capita and the proportion of nuclear households. No significant correlation was found between fertility recovery rate and new COVID-19 cases or emergency duration. When classifying fertility rate fluctuation patterns before and after the emergency into four clusters, variations were noted in the proportion of the elderly population, marriage divorce rate and the number of internet searches related to pregnancy intentions across the clusters.

CONCLUSIONS

No association was found between pregnancy intentions related to the spread of infection, such as the number of new cases and the fertility recovery rate following the first state of emergency. Differences in the patterns of decline and recovery during the pandemic were observed based on population composition and internet searches for infection and pregnancy across different prefectures.

Beyond one-size-fits-all suicide prediction: Studying idiographic associations of risk factors for suicide in a psychiatric sample using ecological momentary assessment

The Interpersonal Psychological Theory of Suicide (IPTS) states that thwarted belongingness (TB), perceived burdensomeness (PB), and hopelessness are risk factors for suicidal ideation. This ecological momentary assessment (EMA) study aimed to (1) demonstrate that there is substantial between-person variability in the association between IPTS predictors and suicidal ideation, (2) identify clusters of patients for which the predictors differently predict suicidal ideation, and (3) examine whether identified clusters are characterized by specific patient characteristics. EMA data were collected ten times per day for six days in 74 psychiatric inpatients and was analyzed with dynamic structural equation modelling. Idiographic associations were obtained and clustered using k-means clustering. We found substantial between-person variability in associations between IPTS predictors and suicidal ideation. Four distinct clusters were identified and different risk factors were relevant for different clusters. In the largest cluster (n = 36), none of the IPTS predictors predicted suicidal ideation. Clusters in which associations between IPTS variables and suicidal ideation were stronger showed higher suicidal ideation, depression, and lower positive affect. These findings suggest that a one-size-fits-all model may not adequately reflect idiosyncratic processes leading to suicidal ideation. A promising avenue might be to use idiographic approaches to personalize prediction and interventions.

The cluster analysis of traditional Chinese medicine authenticity identification technique assisted by chemometrics

This study explore the authenticity identification technique of traditional Chinese medicine (TCM) using chemometrics in conjunction with cluster analysis. A clustering Gaussian mixture model was constructed and applied for the data clustering analysis of four types of TCM. Chemical measurements combined with discrete wavelet transform (DWT), Fourier transform infrared spectroscopy (FTIR), and Fourier self-deconvolution (FSD) were utilized for the detailed differentiation of Bupleurum scorzonerifolium, Bupleurum yinchowense, Bupleurum marginatum, and Bupleurum smithii Wolff var. parvifolium. Differences in the attenuated total reflection-FTIR (ATR-FTIR) spectra among the four TCMs were observed. Utilizing clustering algorithms, the one-dimensional DWT of the infrared spectra of samples was employed for the authentication of Chinese herbal medicines. The model demonstrates optimal performance throughout 2000 rounds of network training. The accuracy (88.6 %), sensitivity (86.5 %), and specificity (82.7 %) of the model constructed in this study significantly surpassed those of the CNN model: accuracy (67.7 %), sensitivity (70.4 %), and specificity (68.5 %) (P < 0.05). By setting the cluster size K = 5 and the number of Gaussian mixture model components to 5, the model effectively fits the actual number of categories within the dataset. Infrared spectroscopy analysis revealed distinct carbon-oxygen stretching vibration absorption peaks between 1025 and 1200 cm-1 for Bupleurum scorzonerifolium, Bupleurum yinchowense, Bupleurum marginatum, and Bupleurum smithii Wolff var. parvifolium, indicating strong absorption peaks of carbohydrates. A comprehensive structural information analysis revealed a similarity of above 0.982 among the four types of TCM. Combined with chemometrics and intelligent algorithm-based cluster analysis, successful and accurate authentication of TCM authenticity was achieved, providing an effective methodology for quality control in TCM.

Modeling the Heterogeneity of Post-Stroke Gait Control in Free-Living Environments Using a Personalized Causal Network

Post-stroke gait control is a complex, often fail to account for the heterogeneity and continuity of gait in existing gait models. Precisely evaluating gait speed adjustability and gait instability in free-living environments is important to understand how individuals with post-stroke gait dysfunction approach diverse environments and contexts. This study aimed to explore individual causal interactions in the free-living gait control of persons with stroke. To this end, fifty persons with stroke wore an accelerometer on the fifth lumbar vertebra (L5) for 24 h in a free-living environment. Individually directed acyclic graphs (DAGs) were generated based on the spatiotemporal gait parameters at contemporaneous and temporal points calculated from the acceleration data. Spectral clustering and Bayesian model comparison were used to characterize the DAGs. Finally, the DAG patterns were interpreted via Bayesian logistic analysis. Spectral clustering identified three optimal clusters from the DAGs. Cluster 1 included persons with moderate stroke who showed high gait asymmetry and gait instability and primarily adjusted gait speed based on cadence. Cluster 2 included individuals with mild stroke who primarily adjusted their gait speed based on step length. Cluster 3 comprised individuals with mild stroke who primarily adjusted their gait speed based on both step length and cadence. These three clusters could be accurately classified based on four variables: Ashman's D for step velocity, Fugl-Meyer Assessment, step time asymmetry, and step length. The diverse DAG patterns of gait control identified suggest the heterogeneity of gait patterns and the functional diversity of persons with stroke. Understanding the theoretical interactions between gait functions will provide a foundation for highly tailored rehabilitation.

Computing advertising intelligent computing and push based on artificial intelligence in the big data era

In the era of big data, intelligent computation and precise targeting of advertisements rely on artificial intelligence (AI) technology to enhance advertising effectiveness and user experience. As advertising media shift from traditional outlets to the internet and mobile devices, manually filtering and targeting advertisements has become increasingly ineffective due to the vast amount of information and user groups. Consequently, AI-based advertisement computation and targeting technologies have emerged. This paper explores how modern technologies and AI applications can be used to achieve intelligent computation, filtering, and targeting of advertisements, thereby improving ad effectiveness and profitability. The study optimizes ad targeting effectiveness and increases return on investment (ROI) by comparing ad campaigns between Group A, which did not use ad targeting algorithms, and Group B, which employed AI-based targeting algorithms. The experimental results show that the average ROI for tourism, shopping, and rental ads in Group A were 153.03 %, 232.32 %, and 192.57 %, respectively, while Group B's average ROI were 173.96 %, 288.74 %, and 216.12 %. These results indicate that AI-based ad targeting algorithms can significantly improve the ROI of ad campaigns compared to non-algorithmic approaches, suggesting that ad targeting algorithms can help advertisers achieve higher profits.