Kinematic Characteristics Research Articles

Decoding communicative action vitality forms in social contexts

The form of an action conveys important information about the agent performing it. Humans may execute the same action in different ways, e.g., vigorously, gently or rudely. This fundamental affective component of the action has been named vitality forms (VFs) by Stern. To date, despite the fundamental role of VFs in social communication, the kinematic features characterizing them have been rarely studied. The aims of the present study are twofold: to investigate spatiotemporal characteristics of transitive and intransitive actions performed with different VFs; to investigate whether and how it is possible to recognize these aspects of action automatically. For this purpose, we asked two actors to perform seven actions with VFs (gentle and rude) or without VFs (neutral, slow and fast). Thousand repetitions of actions were collected, and their kinematics was recorded by using a motion capture system. Twenty-two kinematic features were identified from kinematic data. Results indicate that VFs are not merely characterized by a modulation of a single motion parameter such as velocity or acceleration, but by a combination of different spatiotemporal properties. Finally, we also demonstrate that automatic recognition of VFs is possible using traditional machine learning methods, with an accuracy of 87.3%. Moreover, this recognition is also feasible for action types do not present in the training set, with an accuracy of 74.2%. These results will have significant implications in the future across various fields, including neuroscience, social robotics, and the development of virtual agents. For instance, it could enable artificial agents to recognize human attitudes and adapt their behavior appropriately to the partner during interactions. Moreover, understanding the VFs features could be useful in designing rehabilitative interventions for conditions involving social and communicative impairments, such as autism.

A wing-flapping robot with a bio-inspired folding mechanism derived from the beetle's hind wing.

When the beetle lands on the target, the hind wings fold regularly to form smaller wing packages and are hidden on the ventral side of the elytra due to the interaction between the elytra and abdomen. Its complex folding pattern is attributed to the flexibility of the hind wings, the super-elasticity of the folding joints, and the special geometric morphology of the veins. The corrugation and folding pattern of the hind wings can provide new insights for the design of folding anti-collision mechanisms and the improvement of aerodynamic performance of ornithopter. This paper first proposes a beetle-type ornithopter with foldable wings based on the folding mechanism and kinematic characteristics of the beetle's hind wings. Subsequently, a series of numerical simulations were conducted on flapping wing robot to explore its flapping kinematics, folding stability, structural stiffness. Finally, the force generation of flapping wings was tested on the fabricated prototype.

A sea ice deformation and rotation rate dataset (2017–2023) from the Environment and Climate Change Canada automated sea ice tracking system (ECCC-ASITS)

Abstract. Sea ice forms a thin but horizontally extensive boundary between the ocean and the atmosphere and has complex, crust-like dynamics characterized by intermittent sea ice deformations. The heterogeneity and localization of these sea ice deformations are important characteristics of the sea ice cover that can be used to evaluate the performance of dynamical sea ice models against observations across multiple spatial and temporal scales. Here, we present a new pan-Arctic sea ice deformation and rotation rate (SIDRR; https://doi.org/10.5281/zenodo.13936609, Plante et al., 2024a) dataset derived from the RADARSAT Constellation Mission (RCM) and Sentinel-1 (S1) synthetic aperture radar (SAR) imagery from 1 September 2017 to 31 August 2023. The SIDRR estimates are derived from contour integrals of triangulated ice motion data, obtained from the Environment and Climate Change Canada automated sea ice tracking system (ECCC-ASITS). The SIDRR dataset is not regularized and consists of stacked data from multiple SAR images computed on a range of spatial (4–10 km) and temporal (0.5–6 d) scales. It covers the entire Arctic Ocean and all peripheral seas except the Okhotsk Sea. Uncertainties associated with the propagation of tracking errors on the deformation values are included. We show that rectangular patterns of deformation features are visible when the sampled deformation rates are lower than the propagation error. This limits the meaningful information that can be extracted in areas with low SIDRR values but allows for the study of linear kinematic features with a high SIDRR signal-to-noise ratio. The spatial coverage and range of resolutions of the SIDRR dataset provide an interesting opportunity to investigate regional and seasonal variability in sea ice deformation statistics across scales, and these data can also be used to determine metrics for model evaluation.

Kinematics of the Milky Way from the Statistical Analysis of the Gaia Data Release 3

The aim of the analysis of data from the Gaia Space Observatory is to obtain kinematic parameters of the collective motion of stars in a part of our galaxy. This research is based on a statistical analysis of the motion of 55,038,539 stars selected in different directions from the Sun up to a distance of 3–6 kpc. We developed statistical methods for the analysis working with input data represented by the full astrometric solution (five parameters). Using the proposed statistical methods, we obtained the local velocity of the Sun U⊙,V⊙,W⊙=(9.58,16.25,7.33)±(0.05,0.04,0.02)stat±(0.7,0.9,0.1)syst km/s and the rotation velocity of the galaxy at different radii. For the Sun’s orbit radius, we obtained the velocity of the galaxy rotationVc≈234±4 km/s. Collective rotation slows down in the region under study linearly with distance from the disk plane: ΔV/ΔZ≅33.5kms−1kpc−1. We showed that the different kinematic characteristics and distributions, which depend on the position in the galaxy, can be well described in the studied 3D region by a simple Monte Carlo simulation model, representing an axisymmetric approximation of the galaxy kinematics. The optimal values of the six free parameters were tuned by comparison with the data.

A real-time prediction framework for energy consumption of electric buses using integrated Machine learning algorithms

An accurate prediction of energy consumption in electric buses (EBs) can effectively reduce driving range anxiety and facilitate bus scheduling. Existing studies have not provided real-time predictions based on distance traveled using integrated machine learning methods. This study proposes a framework for predicting EB energy consumption, which is primarily divided into energy consumption estimation, kinematic feature prediction, and energy consumption prediction. The framework begins by fusing high-resolution real-world EB data with weather and road information, from which five types of influencing factors are extracted for different driving distances. An eXtreme Gradient Boosting (XGBoost) model is developed to evaluate feature importance and estimate the energy consumption rate (ECR). The SHapley Additive explanation (SHAP) method is then used to analyze the factors affecting the ECR. To predict important kinematic characteristics, spatial and temporal characteristics are captured using Long Short-Term Memory (LSTM) and a fully connected neural network. Finally, the predicted kinematic characteristics and the XGBoost model are combined to enable real-time prediction of the ECR. The results indicate that estimation and prediction accuracies gradually improve with increased driving distance. The mean absolute error of average ECR decreases from 43.9 % for 100 m to 7.5 % for 16 km. Temperature, bus stop density, and peak periods emerge as the most significant external factors after 8 km. This framework shows an improvement of over 10 % in most scenarios compared with other models in the literature, enabling individual forecasts of energy consumption currently in transit and aiding in the calculation of remaining battery-supported distance.

Simulating seismic wavefields using generative artificial intelligence

Simulating realistic seismic wavefields is crucial for a range of seismic tasks, including acquisition designing, imaging, and inversion. Conventional numerical seismic wave simulators are computationally expensive for large 3D models, and discrepancies between simulated and observed waveforms arise from wave equation selection and input physical parameters such as the subsurface elastic models and the source parameters. To address these challenges, we adopt a data-driven artificial intelligence approach and propose a conditional generative modeling (CGM) framework for seismic wave simulation. The novel CGM framework learns complex 3D wave physics and subsurface heterogeneities from the observed data without relying on explicit physics constraints. As a result, trained CGM-based models act as stochastic wave-propagation operators encoded with a local subsurface model and a local moment tensor solution defined by training data sets. Given these models, we can simulate multicomponent seismic data for arbitrary acquisition settings within the area of the observation, using source and receiver geometries and source parameters as input conditional variables. In this study, we develop four models within the CGM framework — CGM-GM-1D/3D, CGM-Wave, and CGM-FAS — and demonstrate their performance using two seismic data sets: a small low-density data set of natural earthquake waveforms from the San Francisco Bay Area, a region with high seismic risks, and a large high-density data set from induced seismicity records of the Geysers geothermal field. The CGM framework reproduces the waveforms, the spectra, and the kinematic features of the real observations, demonstrating the ability to generate waveforms for arbitrary source locations, receiver locations, and source parameters. We address key challenges, including data density, acquisition geometry, scaling, and generation variability, and we outline future directions for advancing the CGM framework in seismic applications and beyond.

Kinematic characteristics in patients with subcortical vascular cognitive impairment: a quantitative analysis of digitized spiral drawing metrics

Subcortical vascular cognitive impairment (SVCI) is caused by white matter degeneration. Multiple lacunar infarction due to deep small vascular disease around the subcortical structure and incomplete infarction of white matter owing to decreased perfusion are some of the causes of vascular cognitive impairment. Damage to subcortical structures and prefrontal-cortical circuits can affect motor skills along with cognitive abilities. In this study, we identified the indicators, such as pen pressure, speed, and deviation (degree of deviation from the target line), which can be observed in drawing performance in patients with SVCI. A total of 24 patients with SVCI and 26 healthy adults (HC) were subjected to digitized drawing along straight lines and spiral drawing (drawing along spiral lines, free spiral drawing). In the drawing along straight lines, the SVCI group showed a greater deviation than the HC group. In the drawing along spiral lines, the SVCI group showed greater pen pressure and deviation than the HC group. In the free spiral drawing tasks, the SVCI group showed greater pen pressure and deviation than the HC group. In the free spiral drawing task, the deviation of the SVCI group increased as the wheels of the spiral increased in the clockwise and counterclockwise directions. The SVCI group showed greater pressure to hold and press pencils during drawing compared to HC due to damage to cognitive and motor skills; Further, it deviated from the provided target line. These results may be due to complex cognitive-motor deficits, motor patterns, or impaired motor termination and conversion following damage to the frontal-subcortical circuit. This study provides essential clinical basic data that can be utilized to precisely determine the diminished kinematic behavior within the SVCI group.

Kinematic Characteristics of the Non-Throwing Arm During the Completion Phase of the Glide Shot Put in Elite Female Athletes: A Case Study

Background: This study analyzes the biomechanical contributions of the non-throwing arm during the completion phase of the glide shot put technique, focusing on its roles in performance optimization. Methods: Data from a Chinese elite female shot-putter were collected during a national championship, with three-dimensional kinematic analyses and Spearman correlation to assess joint displacement, velocity, and angular changes. Results: Distal joints of the non-throwing arm exhibited greater displacement but lower peak velocity than proximal joints. Angular changes showed a flexion trend in the elbow and shoulder, with brief extension phases in the elbow. During the completion phase, the shoulder velocity of the non-throwing arm positively correlated with shot put velocity (rs = 0.72, p < 0.05) but negatively correlated with the velocity of the elbow (rs = −0.46, p < 0.05), wrist (rs = −0.41, p < 0.05), and center of mass (rs = −0.66, p < 0.05). The elbow velocity positively correlated with shot put velocity (rs = 0.56, p < 0.05) but negatively correlated with velocities of the shoulder (rs = −0.59, p < 0.05), wrist (rs = −0.79, p < 0.05), and center of mass (rs = −0.91, p < 0.05). Wrist velocity exhibited similar correlations. Conclusions: These findings underscore the active role of the non-throwing arm in enhancing shot put performance by influencing the center of mass movement, rotational mechanics, and energy transfer, providing actionable guidance for elite training optimization.

Kinematic characteristics of elastic screen surface and screening performance of material groupsunder different additional excitation conditions

ABSTRACT Screening is indispensable process, which is extensively used in coal separation on a global scale. In traditional screening, material stratification is difficult to achieve, which leads to a poor screening effect. In this study, balls are added under the elastic screen surface to increase the screen amplitude and promote the loose stratification of materials. The screen surface motion characteristics, the changes of the material layer on the screen, the material distribution characteristics under the screen, and the screening efficiency are studied by means of vibration tests, high-speed dynamic recording, and screening tests. The experimental results show that when 2-ball additional excitation is applied, the highest screening efficiency of 79.45% is reached, the elastic screen surface can produce a large amplitude of up to 24.77 mm, the screen surface deformation is larger, the material layer is looser, material migration is faster, and the screening effect is better. The main screening area of the material is the second sections of the screening surface, and the screening efficiency of fine particles is mainly affected by the screening efficiency of complex excitation screening. With the increase of the screening surface length, the screening effect of complex excitation elastic screening becomes better.

Exploring the Feasibility of Head-Tracking Data for Cybersickness Prediction in Virtual Reality

Cybersickness remains a significant barrier to the widespread adoption of virtual reality (VR) technology. Traditional methods for predicting cybersickness rely on self-reported questionnaires or physiological signals from specialized sensors, which have their limitations. This study explores the potential of using real-time, easily acquired head-tracking data (HTD) from standard VR headsets as a scalable alternative for estimating cybersickness. Twelve participants engaged in a VR session using an Oculus Quest 2 headset while their HTD were recorded. Kinematic metrics such as linear and angular velocity, acceleration, and jerk were computed from the HTD, including positional and angular parameters. Participants’ cybersickness levels were assessed using the Virtual Reality Sickness Questionnaire. While exploratory data analysis revealed no significant direct correlation between individual kinematic variables and cybersickness scores, machine learning models were employed to identify predictive patterns. Subsequently, four regression models, including Random Forest, Gradient Boosting, K-Nearest Neighbors, and Support Vector Machines, were trained and evaluated using the computed kinematic features to predict the cybersickness score. Among these, the Gradient Boosting model demonstrated superior performance, accurately predicting cybersickness scores with normalized differences less than 3.08% on unseen data. This approach offers a scalable and practical solution for real-time cybersickness prediction in VR applications and compliments other techniques that rely on physiological sensors, hardware, or user profiles.

Circuit Modules for Flexible Locomotion.

Locomotion, like all behaviors, possesses an inherent flexibility that allows for the scaling of movement kinematic features, such as speed and vigor, in response to an ever-changing external world and internal drives. This flexibility is embedded in the organization of the spinal locomotor circuits, which encode and decode commands from the brainstem and proprioceptive feedback. This review highlights our current understanding of the modular organization of these locomotor circuits and how this modularity endows them with intrinsic mechanisms to adjust speed and vigor, thereby contributing to the flexibility of locomotor movements.

Which is more effective in thawing frozen rooster sperm: varying temperature or duration?

Cryopreservation of poultry sperm is crucial for preserving genetic diversity and protecting endangered breeds. Rooster sperm is highly sensitive to cryopreservation due to its high polyunsaturated fatty acid content, making it prone to damage during freezing and thawing. This study evaluated the effects of thawing temperatures and storage conditions on sperm quality, including motility, morphology, and viability. Frozen rooster semen samples were thawed at 37°C for 30 seconds, 60°C for 5 seconds, or 72°C for 5 seconds and stored at 4°C for up to 48 hours. Sperm quality parameters, including motility, kinematic characteristics, abnormal morphology, and viability, were assessed at 0, 3, 6, 9, 12, 24, and 48 hours using a Computer-Assisted Semen Analyzer (CASA). Post-thaw motility varied significantly between thawing temperatures at 24 and 48 hours (p < 0.05). Progressive and rapid progressive motility also differed significantly at 24 hours (p < 0.05). Sperm viability showed statistical differences across thawing groups at 24 and 48 hours (p < 0.05), while morphological abnormalities were significant at 12 and 48 hours (p < 0.05). Across all groups, sperm quality parameters varied significantly at each time point (p < 0.05). Thawing at 37°C and storing at 4°C for up to 24 hours optimizes sperm motility and viability, minimizing cryodamage and ensuring functional preservation. This approach is effective for short-term storage and crucial for sustaining genetic diversity and fertility in poultry breeding programs.

Simulation Analysis of Powered Parachute Motion Characteristics based on Motor Thrust and Flap Deflection Control

In this paper, a distributed new type of parafoil is proposed, conducting its dynamic modeling and motion characteristic analysis, thereby laying the foundation for subsequent trajectory tracking research and physical experimentation. First, the dynamics of the parafoil and payload are analyzed, and the dynamic model of the entire parafoil system is derived from this analysis. Then, the kinematic characteristics under control inputs of bilateral pull-down, unilateral pull-down, and unilateral pull-down with differential thrust are examined using the simulation model; Simulation results indicate that the established 8-DOF dynamic model accurately describes the parafoil's motion. The proposed differential control increases yaw motion response speed and provides a basis for subsequent trajectory tracking research and physical experiments. The proposed differential control enhances the yaw motion response speed and reduces the relative motion between the wing parachute and payload to a certain extent. Simulation analysis verifies the model's validity and the control method's feasibility.

Kinematic and kinetic characteristics of individuals with coping lateral ankle sprain during landing and walking/running tasks: A systematic review with meta-analysis.

We aimed to synthesize the kinematics and kinetics during landing and walking/running tasks of ankle copers compared with patients with chronic ankle instability and controls. We systematically searched PubMed, CINAHL, SPORTDiscus, and Web of Science. Tri-planar lower extremity biomechanics (joint angle and moment at maximum and initial contact, and joint displacement) were synthesized using standard mean difference and 95% confidence intervals. We assessed methodological quality, study heterogeneity, and publication bias. We identified 13 studies. Pooled evidence indicated that copers had less inversion and eversion angles, and ankle frontal displacement during landing compared with patients with chronic ankle instability (|standard mean difference| 0.31-0.40). Copers revealed greater dorsiflexion, plantarflexion, and ankle sagittal displacement during landing and less plantarflexion moment during walking than patients with chronic ankle instability (|standard mean difference| 0.29-0.46); however, most ankle biomechanics of copers were not significantly different from those of controls. In the proximal joints, copers revealed less knee valgus angle at maximum and initial contact (|standard mean difference| 0.25-0.33); and less hip flexion angle at maximum and initial contact, hip external rotation angle and sagittal and frontal displacement during landing (standard mean difference 0.49-0.73) than patients with chronic ankle instability. Copers had less hip extension moment during walking and greater maximum hip adduction angle during landing than controls (|standard mean difference| 0.66-0.77). Our results provide insight into the lower extremity biomechanics of copers during dynamic tasks, contributing injury coping mechanisms and developing rehabilitation programs to return to sports.

Lateral End-Range Movement Profile and Shot Effectiveness During Grand Slam Tennis Match-Play.

End-range movements are among the most demanding but least understood in the sport of tennis. Using male Hawk-Eye data from match-play during the 2021-2023 Australian Open tournaments, we evaluated the speed, deceleration, acceleration, and shot quality characteristics of these types of movement in men's Grand Slam tennis. Lateral end-range movements that incorporated a change of direction (CoD) were identified for analysis using k-means (end-range) and random forest (CoD) machine learning models. Peak speed, average deceleration into the CoD, average reacceleration out of the CoD, and the quality of the shot played were computed. Players were grouped based on their ATP rankings (top 10, top 50, and outside top 50) to examine the influence of ranking on movement profiles and shot effectiveness. Our data showed that end-range movements profiles of top 10 and top 50 players were characterized by higher peak speed (d=0.3-0.88), deceleration intensity (d=0.25-0.63), and acceleration intensity (d=0.06-0.51) when compared to players outside the top 50 (p<0.05). Top 10 players also demonstrated greater peak speeds (d=0.59) and acceleration intensities (d=0.45) compared to top 50 players (p<0.05). There was a nonlinear inverse relationship between peak speed and shot quality, such that, as peak speed increased, shot quality decreased-notwithstanding that top 10 players were more likely to hit high-quality shots at higher peak speeds. These results quantify the discrete kinematic characteristics of the sport's most challenging movement sequence and reveal, for the first time, that higher ranked players may possess superior movement potential on court.

Течение ньютоновской жидкости в смесителях различных конфигураций

The flow of a viscous incompressible fluid in agitators with solid and anchor blades is analyzed. Mathematical formulation for the problem involves Navier-Stokes and continuity equations in the plane approximation. A solution algorithm is developed based on the control volume method and a SIMPLE correction procedure. The differential equations are discretized with unstructured triangular meshes that take into account the geometric features of the flow region. Tests were performed to verify the approximation convergence and to estimate the order of accuracy of the numerical scheme, as well as to verify the original calculation program. Parametric studies were carried out for Reynolds numbers in range of 0.1-100, which is characteristic of the technology of fluid materials processing in industrial mixers. Distributions of the kinematic and dynamic characteristics of the flow were obtained, and the flow patterns, whose specific feature is the presence of circulation zones in paddle agitators of different configurations, were demonstrated. Calculations are done to determine the position of marker particles and the evolution of reference lines, which give a visual representation of the process and demonstrate the presence of areas of uniform and non-uniform mixing. A quantitative mixing quality parameter is introduced to compare agitators of different configurations with each other and to consider the process of mixing over time. The process is evaluated quantitatively by the values of the integral of the dissipative function, which shows the energy consumption, and by the value of the quantitative parameter of the inhomogeneity of the marker particle distribution. The latter allows a more detailed study of the mixing process over the volume over time.

Deep Temporal Clustering of Pathological Gait Patterns in Post-Stroke Patients Using Joint Angle Trajectories: A Cross-Sectional Study.

Rehabilitation of gait function in post-stroke hemiplegic patients is critical for improving mobility and quality of life, requiring a comprehensive understanding of individual gait patterns. Previous studies on gait analysis using unsupervised clustering often involve manual feature extraction, which introduces limitations such as low accuracy, low consistency, and potential bias due to human intervention. This cross-sectional study aimed to identify and cluster gait patterns using an end-to-end deep learning approach that autonomously extracts features from joint angle trajectories for a gait cycle, minimizing human intervention. A total of 74 sub-acute post-stroke hemiplegic patients with lower limb impairments were included in the analysis. The dataset comprised 219 sagittal plane joint angle and angular velocity trajectories from the hip, knee, and ankle joints during gait cycles. Deep temporal clustering was employed to cluster them in an end-to-end manner by simultaneously optimizing feature extraction and clustering, with hyperparameter tuning tailored for kinematic gait cycle data. Through this method, six optimal clusters were selected with a silhouette score of 0.2831, which is a relatively higher value compared to other clustering algorithms. To clarify the characteristics of the selected groups, in-depth statistics of spatiotemporal, kinematic, and clinical features are presented in the results. The results demonstrate the effectiveness of end-to-end deep learning-based clustering, yielding significant performance improvements without the need for manual feature extraction. While this study primarily utilizes sagittal plane data, future analysis incorporating coronal and transverse planes as well as muscle activity and gait symmetry could provide a more comprehensive understanding of gait patterns.

Association of real life postural transitions kinematics with fatigue in neurodegenerative and immune diseases

Fatigue is prevalent in immune-mediated inflammatory and neurodegenerative diseases, yet its assessment relies largely on patient-reported outcomes, which capture perception but not fluctuations over time. Wearable sensors, like inertial measurement units (IMUs), offer a way to monitor daily activities and evaluate functional capacity. This study investigates the relationship between sit-to-stand and stand-to-sit transitions and self-reported physical and mental fatigue in participants with Parkinson’s, Huntington’s, rheumatoid arthritis, systemic lupus erythematosus, primary Sjögren’s syndrome and inflammatory bowel disease. Over 4 weeks, participants wore an IMU and reported fatigue levels four times daily. Using mixed-effects models, associations were identified between fatigue and specific kinematic features, such as 5th and 95th percentiles of sit-to-stand performance, suggesting that fatigue alters the control and effort of movement. These kinematic features show promise as indicators for fatigue in these patient populations.

Kinematic Characteristics and Water Mass Transports of Submesoscale Coherent Vortices in the Northeastern South China Sea

AbstractAs a unique phenomenon occurring in the subsurface ocean, submesoscale coherent vortices (SCVs) are believed to have a pivotal role in the long‐distance ocean tracer transports. SCVs have been widely observed in the global oceans; however, most of them are captured accidentally, and their kinematic characteristics and water mass transports have only been studied in a limited number of regions. Here, we use 4‐year observations of velocity, temperature, and salinity from five moorings in the northeastern South China Sea (NESCS) to examine dozens of newly discovered SCVs. A total of 34 SCVs were identified during the observational period, including 25 convex lens‐like anticyclones and 9 concave lens‐like cyclones. The maximum swirl velocity, mean radius, and vertical length scale of the anticyclones (cyclones) are 0.19 ± 0.07 m s−1 (0.19 ± 0.07 m s−1), 26.4 ± 13.9 km (17.0 ± 5.4 km), and 204 ± 62 m (188 ± 53 m), respectively. Vertically, the velocity structure of the observed SCVs conforms to a Gaussian function when the effect of stratification is removed. Water mass analysis suggests that 88% (30/34) of the SCVs carried Kuroshio water, which demonstrates the mechanism proposed by Zhang et al. (2022), https://doi.org/10.1029/2021jc018117 that they are formed by Kuroshio‐islands interactions in the Luzon Strait. This category of SCVs is therefore named Luzon Strait island wake eddies (Liddies). We further estimate that Liddies can result in an equivalent mean volume transport of 0.20 Sv westward across the Luzon Strait, which suggests that they play a non‐negligible role in the subsurface water transports between the NESCS and the northwestern Pacific.

A novel feature extraction method based on dynamic handwriting for Parkinson's disease detection.

Parkinson's disease (PD) is a common disease of the elderly. Given the easy accessibility of handwriting samples, many researchers have proposed handwriting-based detection methods for Parkinson's disease. Extracting more discriminative features from handwriting is an important step. Although many features have been proposed in previous researches, the insight analysis of the combination of handwriting's kinematic, pressure, and angle dynamic features is lacking. Moreover, most existing feature is incompletely represented, with feature information lost. Therefore, to solve the above problems, a new feature extraction approach for PD detection is proposed using handwriting. First, built on the kinematic, pressure, and angle dynamic features, we propose a moment feature by composed these three types of features, an overall representation of these three types of features information. Then, we proposed a feature extraction method to extract time-frequency-based statistical (TF-ST) features from dynamic handwriting features in terms of their temporal and frequency characteristics. Finally, we proposed an escape Coati Optimization Algorithm (eCOA) for global optimization to enhance classification performance. Self-constructed and public datasets are used to verify the proposed method's effectiveness respectively. The experimental results showed an accuracy of 97.95% and 98.67%, a sensitivity of 98.15% (average) and 97.78%, a specificity of 99.17% (average) and 100%, and an AUC of 98.66% (average) and 98.89%. The code is available at https://github.com/dreamhcy/MLforPD.