华沙医科大学

Machine-learning representations of gait waveforms are increasingly used in movement analysis, but their added value over conventional biomechanical descriptors remains uncertain, particularly in pediatric cohorts. We evaluated whether implicit neural representations (INRs) provide useful encodings of pediatric gait waveforms for reconstruction and developmental modeling. The full cohort comprised 78 healthy children and adolescents, of whom 73 had waveform files conforming to the predefined data structure required for the repeated cross-validation analyses. Selected waveforms were modeled using several INR architectures, and reconstruction quality was assessed using R2, RMSE, and MAE. For the INR feature-ablation analyses, one architecture was assigned to each waveform class based exclusively on reconstruction performance, independently of the age targets. Side-specific, bilateral-asymmetry, signal-specific, and descriptor-family feature sets were evaluated using ten repeats of five-fold cross-validation. A separate matched benchmark in the same 73 participants compared compact Fourier-MLP-derived features with conventional biomechanical descriptors and a fold-wise raw-waveform PCA baseline. In the INR ablation analyses, combining side-specific and bilateral-asymmetry features achieved R2=0.700±0.017 for chronological-age regression, and Energy/AUC descriptors were the strongest individual feature family (R2=0.696±0.031). For exploratory three-class age-group classification (4–7, 8–12, and 13–18 years), the combined INR representation reached a balanced accuracy of 0.749 ± 0.034. In the matched representation benchmark, configurations containing conventional biomechanical descriptors performed best. The biomechanics-only baseline achieved regression R2=0.837±0.070 and balanced accuracy =0.798±0.109, whereas biomechanics+PCA achieved the highest mean regression performance (R2=0.842±0.066); this difference was not statistically significant. Adding INR-derived features did not improve regression performance and significantly reduced the evaluated classification metrics. INRs should therefore be considered complementary continuous waveform representations rather than replacements for established biomechanical descriptors.

AIMS:To describe and quantify sex differences in the screening electrocardiograms (ECGs) of athletes. METHODS AND RESULTS:Five databases (MEDLINE, EMBASE, Scopus, SPORTDiscus, and Web of Science) were searched from inception until October 2024. Included studies were original research articles examining athletes aged 16-40 years, who had a 12-lead screening ECG, and where analysis was stratified by sex. Risk of bias was assessed using a validated tool. A meta-analysis was performed, using a random-effects model, assessing proportions of athlete normal, borderline, and abnormal ECG features (according to the 2017 International Criteria) and several ECG measurements. Eighty-five cross-sectional studies were included. The studies comprised 19 069 female athletes (mean age 20.6 ± 2.8 years) and 57 745 male athletes (mean age 21.3 ± 3.1 years). Female athletes were more likely to have abnormal T-wave inversion (TWI) (OR 2.3, 95% CI: 1.5-3.5), and isolated TWIV1-V3 (OR 5.6, 95% CI: 3.2-9.9) compared with males. No female athletes and two male athletes with isolated TWIV1-V3 were diagnosed with a condition associated with sudden cardiac arrest or death (both male athletes diagnosed with arrhythmogenic cardiomyopathy). Female athletes were also more likely to have an abnormal ECG per the International Criteria, though the finding was not statistically significant (OR 1.3, 95% CI: 0.7-2.4). Female athletes had a 16 ms (95% CI: 4-27 ms) longer QTc interval than male athletes. CONCLUSION:Compared with male athletes, female athletes were twice as likely to have abnormal TWI and six times as likely to have isolated TWIV1-V3, a finding which was not accompanied by diagnoses of cardiac pathology in female athletes. These data should help inform sex-specific aspects of athlete ECG screening guidelines. LAY SUMMARY:This paper describes sex differences in athlete screening electrocardiograms. KEY FINDINGS:

The development of efficient thermoelectric materials for direct waste heat–to–electricity conversion remains a major challenge, particularly for high-temperature applications. In this paper, a systematic first-principles examination of the structural, electronic, elastic, thermal and thermoelectric characteristics of double perovskite oxides, Ba2XReO6 (X = Li, Rb, Cs), is conducted using the density functional theory, alongside Boltzmann transport theory. Structural optimization proves that all compounds are thermodynamically stable in the cubic Fm3̅m phase with negative formation energies. Furthermore, ab initio molecular dynamics simulations performed at 300 K confirm the dynamical stability of the optimized structures. The electronic structure analysis reveals semiconducting behavior in Ba2CsReO6, Ba2LiReO6, and Ba2RbReO6, all exhibiting indirect and narrow band gaps, which are ideal characteristics for thermoelectric transport. The elastic constants obtained and used have met the Born stability requirements, indicating mechanical stability and ductile behavior. A systematic decrease in Debye temperature, melting temperature, and lattice thermal conductivity with increasing A-site ionic radius led to ultralow lattice thermal conductivity of Ba2RbReO6 and Ba2CsReO6. Thermoelectric transport calculations show positive Seebeck coefficients and enhanced power factors at elevated temperatures. Of the investigated compounds, the Ba2RbReO6 has the largest thermoelectric figure of merit (ZT), owing to an optimum balance between electrical conductivity and reduced thermal conductivity. Overall, our results suggest that Ba2XReO2, particularly Ba2RbReO2, are promising candidates for high-temperature thermoelectric applications.

Although therapeutic drug monitoring (TDM) is recommended for personalized clozapine (CLO) dose titration, there are no widely accepted international guidelines regarding routine TDM during long-term outpatient CLO treatment and many clinicians have limited access to TDM. This retrospective, naturalistic study aimed to examine whether CLO and norclozapine (NCLO) concentrations measured only at the time of inpatient care could predict the risk of psychiatric rehospitalization after discharge, irrespective of the potential influence of confounding factors on long-term outcomes. A total of 141 blood samples from psychiatric inpatients (71 females and 70 males, aged from 18 to 74 years) were analyzed. Serum CLO and NCLO levels were determined with high-performance liquid chromatography coupled with a UV detector. We assessed the rates of rehospitalization over 360 days after discharge. There was no correlation between either CLO or NCLO concentrations and psychiatric rehospitalization rates at any follow-up time points (day 90, 180, or 360). However, patients receiving CLO as a single psychotropic drug had significantly fewer rehospitalizations (assessed at 180 and 360 days of follow-up, p < 0.001) occurred among patients receiving CLO as a single psychotropic drug compared to combination therapy. The risk of rehospitalization was also higher in patients with a greater number of previous hospitalizations (OR = 1.15, p = 0.002) and in those taking additional psychotropic drugs (OR = 1.51, p = 0.010). Our naturalistic retrospective study (based on an in-hospital TDM database) showed no association between the baseline steady-state CLO and NCLO levels established during hospitalization and the risk of psychiatric rehospitalization. However, the use of CLO in combination with other psychotropic drugs and the number of previous hospitalizations correlated with the risk of rehospitalization. Not applicable.

Identification of adolescent e-cigarette use could inform prevention and intervention programming and reduce associated consequences. One way to predict those engaging in use is by examining social media profiles and metrics. Most studies examining substance use content on social media employ self-report or human coding that have methodological limitations. Thus, the current study developed a supervised machine learning algorithm to classify participants into e-cigarette use categories based on Instagram metrics. Participants (n = 67, Mage = 18.27; 64.2