Serkan Örücü

In Particle Swarm Optimization (PSO) and Quantum Inspired PSO (QI-PSO), the initialization source can shape early population diversity, search coverage, convergence behavior, and the later response of candidate generation operators. This effect becomes important in both continuous optimization, where particles move in a real-valued search space, and wrapper feature selection, where candidate solutions are evaluated as binary feature masks through classifier performance. Dragon curves offer another way to define structured search sources because their ordered point sequences can be mapped into a normalized search domain and reused as geometric guides. Based on this idea, this study examines whether random initialization, Sobol initialization, QI candidate generation, and dragon-curve-based geometric guidance change method behavior across continuous optimization and wrapper feature selection. The PSO variant set included Random-PSO, Sobol-PSO, Heighway-PSO, Twindragon-PSO, and Terdragon-PSO, while the QI-PSO variant set included Random-QI-PSO, Sobol-QI-PSO, Heighway-QI-PSO, Twindragon-QI-PSO, and Terdragon-QI-PSO. Heighway, Twindragon, and Terdragon curves were used as ordered geometric sources for particle initialization and stagnation-triggered perturbation. The methods were tested on CEC 2022 benchmark functions in 10-dimensional and 20-dimensional settings and on the UCI Cleveland Heart Disease dataset through wrapper feature selection with KNN-based fitness and additional SVM evaluation. In CEC 2022, the PSO variants produced lower average ranks and lower average mean fitness values than the QI-PSO variants. Random-PSO achieved the lowest average rank and runtime, whereas Terdragon-PSO produced the lowest average mean fitness and the lowest mean best fitness trajectory among the PSO variants. Dragon-guided PSO variants produced more best mean fitness records, but with higher computational cost. In the Cleveland task, Sobol-QI-PSO produced the highest observed mean balanced accuracy among the wrapper FS methods under both KNN and SVM. However, the Holm-corrected comparisons indicated statistically similar balanced accuracy across the compared wrapper methods. Terdragon-PSO produced the highest feature subset stability, while Random-PSO produced the most compact subsets. These results show that initialization, QI candidate generation, and dragon-based guidance vary with problem representation and evaluation dimension. Future studies should evaluate new metaheuristic components through final fitness, runtime, classifier response, selected feature count, selection frequency, and feature subset stability rather than through one performance measure.

Feature selection and hyperparameter optimization are widely used in heart disease prediction, but comparative findings are often affected by differences in preprocessing, search space design, and evaluation budget. This study compared quantum-inspired particle swarm optimization and evolutionary algorithms with their binary variants, namely QI-PSO, QI-EA, QI-BPSO, and QI-BEA, under the same evaluation protocol for joint feature selection and discrete hyperparameter tuning on the UCI Cleveland Heart Disease dataset. The feature mask and classifier hyperparameters were optimized simultaneously in a unified search space using the same preprocessing scheme, inner-validation protocol, and evaluation budget. Decision Tree (DT), Logistic Regression (LR), Support Vector Machine (SVM), and K-Nearest Neighbors (K-NN) classifiers were evaluated using a fixed train-test split. Held-out test performance was reported with 95% confidence intervals. In addition, a no-feature selection baseline was evaluated for each classifier using the final hyperparameter configuration of the best selected feature QI model for the corresponding classifier while retaining all 13 input variables. Precision-recall curves, threshold sensitivity profiles, outer-trial CV Accuracy, runtime, selected feature counts, convergence behavior, and cost-performance distributions were also analyzed. Statistical comparisons of outer-trial distributions were performed using Kruskal-Wallis tests followed by Holm-adjusted Mann-Whitney U post-hoc comparisons. The held-out test results showed classifier-dependent differences among the QI variants. For DT and K-NN, the highest selected-feature test Accuracy was obtained with QI-PSO, whereas for SVM it was obtained with QI-BPSO. In the K-NN setting, QI-BEA, defined in this study as the modified binary implementation of the QI-EA update strategy, produced higher selected-feature Accuracy, F1-score, and MCC than QI-EA. For DT, QI-BPSO yielded the highest ROC-AUC, while for LR all four QI methods produced identical held-out test metrics. Compared with the corresponding no-feature selection baselines, the selected feature QI models showed a clear held-out advantage for K-NN. In DT, the no-feature selection baselines matched the best selected feature accuracy, whereas QI-BPSO yielded the highest DT ROC-AUC. In LR and SVM, the no-feature selection baselines produced stronger held-out results than the selected feature configurations. Precision-recall and threshold sensitivity analyses further showed classifier-dependent overlap and separation patterns beyond single point estimates. At the outer-trial level, QI-BPSO occupied the lowest runtime range, while QI-BEA yielded the highest median CV accuracy for DT, K-NN, and LR. These findings show that the relative behavior of QI variants depends on the classifier family and on whether emphasis is placed on held-out test performance, operating-threshold behavior, inner-loop validation level, or computational cost.

Accurate assessment of joint range of motion (ROM) is essential in both clinical and athletic contexts for monitoring mobility, guiding rehabilitation, and optimizing performance. Traditional tools such as goniometers are widely used but limited by operator dependency and static measurement constraints. Depth-sensing technologies offer a markerless alternative that may enhance practicality in field-based settings. This study examined the validity and reliability of the Microsoft Azure Kinect system compared with a digital goniometer for assessing hip and knee ROM in 30 elite female weightlifters from the Turkish National Team. Participants performed hip abduction, adduction, flexion, extension, and knee flexion, each measured using both tools. Agreement between systems was assessed using Pearson’s correlations, intraclass correlation coefficients [ICC (3,1)], and Bland-Altman analyses. No significant differences were found between methods (p > 0.05), with strong correlations (r = 0.974–0.997) and high intraclass correlation coefficients (ICC = 0.994–0.998), indicating excellent consistency. Bland-Altman analyses showed minimal bias and narrow limits of agreement. These findings support the Azure Kinect as a valid and reliable tool for lower-extremity ROM assessment and a viable alternative to traditional goniometers in field applications.

Kinect is a markerless, portable, and affordable motion analysis tool used in clinical, rehabilitation, and sports settings. This study aimed to assess upper limb (shoulder and elbow) and cervical joint angles in elite female weightlifters using Azure Kinect and a digital goniometer, and to evaluate the Kinect’s validity and reliability. Joint angles were measured in elite female weightlifters (n = 21) using both a digital goniometer and Azure Kinect, with three repetitions per movement under standardized conditions (within a stabilization cage). Mean values were used for analysis. Statistical analysis included descriptive metrics and non-parametric tests. Tool comparisons were conducted using the Wilcoxon signed-rank test, Spearman’s correlation coefficient, and Bland–Altman plots. Reliability was evaluated through the Intraclass Correlation Coefficient (ICC), Coefficient of Variation (CV), and Coefficient of Repeatability (CR), with statistical significance set at p < 0.05. Most joint angle measurements showed no statistically significant differences between the digital goniometer and Azure Kinect (p > 0.05), although small discrepancies appeared in some movements. High correlation coefficients (r = 0.82–0.99) and strong agreement based on Intraclass Correlation Coefficient (ICC = 0.97–0.99) were observed. Bland–Altman analysis revealed minimal systematic bias and narrow confidence intervals. Comparable values in Coefficient of Variation (CV) and Coefficient of Repeatability (CR) indicated high stability and repeatability for both tools. Overall, results support Azure Kinect as a valid and reliable alternative to traditional digital goniometry for assessing cervical and upper extremity joint range of motion.

Microsoft Kinect is an increasingly common tool for low-cost and portable measurement of human movements. The aim of this study was to investigate the potential reliability and reproducibility of Microsoft Kinect Azure in the sit-and-reach test conducted under standardized test conditions in a controlled training environment with elite female weightlifters. The sit-and-reach test was performed with Microsoft Kinect Azure and manual methods and also the reliability and validity were analyzed by comparing the wrist, elbow, shoulder and trunk joint angle measurements with the digital goniometer. Twenty-one healthy female weightlifter athletes who won medals in international championships were included in the study. Sit-and-reach testing was performed using Microsoft Kinect Azure and manual methods in accordance with the standards of the American College of Sports Medicine. The flexion range of motions of each athlete’s wrist, elbow, shoulder and trunk joint angles were measured with a digital goniometer and Microsoft Kinect Azure. Statistical analyses were performed using IBM SPSS 25.0 and Python 3.9.12 software. Descriptive statistics were calculated for all variables, and differences between Kinect and manual methods were evaluated with the Mann-Whitney U test and Wilcoxon Signed-Rank Test. Linear relationships between measurements were analyzed with Spearman correlation coefficients. Intraclass Correlation Coefficient was calculated for the concordance analysis and presented with Bland-Altman plots. The level of significance was accepted as p < 0.05. The findings suggest that Kinect Azure may provide good reproducibility and agreement with manual goniometric measurements for joint angles under the specific structured conditions tested. For parameters such as the elbow, shoulder, wrist, trunk angles and distance traveled, high ICC values (> 0.90) indicated strong consistency between the methods. Kinect Azure may serve as a reliable alternative to manual measurement methods under standardized testing conditions. Further studies are recommended to examine its generalizability to other populations and test environments.

Anthropometric measurements are important in a wide range of areas, from athlete selection to performance analysis and medical exercise applications. With its portable and cost-effective structure, Microsoft Kinect offers significant advantages in measuring human movements and provides valuable data in the fields of sports science and medicine. The aim of this study is to compare certain anthropometric measurements of adolescent female weightlifters in the U-15 age category with Kinect V2 and traditional methods. Twelve athletes who won medals in international weightlifting championships were included in the study. Anthropometric evaluations of each athlete were performed using Kinect V2 and the traditional method. Differences between measurements were analyzed with Bland–Altman plots and Pearson correlation coefficients. Kinect was found to exhibit less variability and higher coherency in measurements of humerus, forearm, trochanter-tibiale laterale, tibial lengths, and shoulder width. It has been observed that the traditional method provides more coherence results in hand length measurements. Kinect was found to exhibit less variability and higher coherency in measurements of forearm, trochanter-tibiale laterale, tibial lengths, and shoulder width, indicating its potential as a reliable tool for these parameters. Notably, Kinect demonstrated high reliability for tibial length (ICC 0.918) and moderate reliability for trochanter-tibiale laterale (ICC 0.737), showing its effectiveness in these measurements. Additionally, Kinect exhibited lower variation and higher coherency in most measurements compared to manual methods, supporting its consistency and repeatability in anthropometric assessments. These results indicate that Kinect may be a suitable tool for some measurements but that traditional methods may be preferable for hand length measurements. These findings suggest that Kinect can be used effectively for certain anthropometric measurements in sports and medical science. Graphical Abstract

The importance of anthropometric measurements for monitoring the physical development of athletes and optimizing training programs is well known. Considering the limitations of traditional methods in terms of accuracy and consistency, the aim of this study is to investigate the potential of Kinect V2 as an alternative. This study wasperformed on 12 youth female weightlifters who won medals in the Youth European Weightlifting Championship. Humerus length, forearm length, hand length, trochanter-tibiale laterale length, tibial length and shoulder width measurements were performed with both Kinect V2 and manual methods. Statistical differences between groups were evaluated using the Mann–Whitney U test. The consistencies of the measurement methods were analyzed with Bland-Altman plots, correlations were determined with Pearson correlation coefficients, and reliability were evaluated with Intraclass Correlation Coefficient (ICC) values. Kinect V2 provided accuracy and consistency comparable to manual methods in measurements of humerus (ICC = 0.532), forearm (ICC = 0.600), and hand length (ICC = 0.760). While medium-level concordance was observed in trochanter-tibiale lateral length measurements (ICC = 0.749), high-level concordance and reliability were found in tibial length (ICC = 0.914) and shoulder width (ICC = 0.869) measurements. There were significant differences between the results of humerus length, forearm length and trochanter-tibiale laterale length obtained with both measurement methods. Therefore, measurements of these parameters require significant care. Findings of this study suggest that Kinect V2 can be a reliable tool for rapid and practical anthropometric assessments in sports settings, but highlight the importance of careful calibration and adjustments for specific measurements. Future studies should examine the use of this device more comprehensively across different sports and populations. Anthropometric measurements are important for monitoring the physical development of athletes and optimizing their training programs. Considering the limitations of traditional methods in terms of accuracy and consistency. Kinect analyzes the movements of athletes during sports education and training processes, making it possible to organize training programs according to individual needs. Kinect V2, which has a high potential for use, especially in medical exercise and clinical applications, also finds a wide range of applications in sports sciences. In this study, a comparative analysis of certain lower and upper extremity anthropometric measurements such as humerus length, forearm length, hand length, trochanter-tibiale laterale length, tibial length and shoulder width of female weightlifting athletes in the youth age category were performed using Kinect V2 and manual methods.

Vücut geliştirme gibi izotonik hareketlerin yapıldığı sporlarda ve rehabilitasyon süreçlerinde, antrenmanların eşzamanlı izlenmesi, yanlış yapılan hareketlerin anında düzeltilebilmesi antrenman yeterliliği ve veriminin belirlenebilmesi bireylerin yaralanma risklerinden uzak ve sağlıklı antrenman yapabilmeleri açısından hayati önem taşımaktadır. Çalışmamızda, bu amaca yönelik hareket modeline dayalı kural tabanlı bir Uzman Sistem (ES) tasarlanmıştır. Tasarlanan sistem, antrenör üzerinden elde edilen LR (Lateral Raise) antrenmanı verileri aracılığıyla, DTW metodu ile karşılaştırmalı olarak test edilmiştir. Test sonuçlarında, tasarladığımız ES'in aldığı kararların DTW metodundan elde edilen kararlar ile karşılaştırmasında %64 oranında doğrulukla, daha yüksek oranda doğruluğa sahip olduğu görülmüştür.

In sports and rehabilitation processes where isotonic movements such as bodybuilding are performed, it is vital for individuals to be able to correct the wrong movements instantly by monitoring the trainings simultaneously, and to be able to train healthily and away from the risks of injury. For this purpose, we designed a new real-time athlete support system using Kinect V2 and Expert System. Lateral raise (LR) and dumbbell shoulder press (DSP) movements were selected as examples to be modeled in the system. Kinect V2 was used to obtain angle and distance changes in the shoulder, elbow, wrist, hip, knee, and ankle during movements in these movement models designed. For the rule base of Expert System developed according to these models, a 2⁸-state rule table was designed, and 12 main rules were determined that could be used for both actions. In the sample trainings, it was observed that the decisions made by the system had 89% accuracy in DSP training and 82% accuracy in LR training. In addition, the developed system has been tested by 10 participants (25.8 ± 5.47 years; 74.69 ± 14.81 kg; 173.5 ± 9.52 cm) in DSP and LR training for four weeks. At the end of this period and according to the results of paired t-test analysis (p < 0.05) starting from the first week, it was observed that the participants trained more accurately and that they enhanced their motions by 58.08 ± 11.32% in LR training and 54.84 ± 12.72% in DSP training.

Monitoring of training performance and physical activity has become indispensable these days for athletes. Wireless technologies have started to be widely used in the monitoring of muscle activation, in the sport performance of athletes, and in the examination of training efficiency. The monitorability of performance simultaneously in the process of training is especially a necessity for athletes at the beginner level to carry out healthy training in sports like weightlifting and bodybuilding. For this purpose, a new system consisting of 4 channel wireless wearable SEMG circuit and analysis software has been proposed to detect dynamic muscle contractions and to be used in real-time training performance monitoring and analysis. The analysis software, the Haar wavelet filter with threshold cutting, can provide performance analysis by using the methods of moving RMS and %MVC. The validity of the data obtained from the system was investigated and compared with a biomedical system. In this comparison, 90.95% ± 3.35 for left biceps brachii (BB) and 90.75% ± 3.75 for right BB were obtained. The output of the power and %MVC analysis of the system was tested during the training of the participants at the gym, and the training efficiency was measured as 96.87% ± 2.74.