Incremental Training Research Articles

Replay-Based Incremental Learning Framework for Gesture Recognition Overcoming the Time-Varying Characteristics of sEMG Signals

Gesture recognition techniques based on surface electromyography (sEMG) signals face instability problems caused by electrode displacement and the time-varying characteristics of the signals in cross-time applications. This study proposes an incremental learning framework based on densely connected convolutional networks (DenseNet) to capture non-synchronous data features and overcome catastrophic forgetting by constructing replay datasets that store data with different time spans and jointly participate in model training. The results show that, after multiple increments, the framework achieves an average recognition rate of 96.5% from eight subjects, which is significantly better than that of cross-day analysis. The density-based spatial clustering of applications with noise (DBSCAN) algorithm is utilized to select representative samples to update the replayed dataset, achieving a 93.7% recognition rate with fewer samples, which is better than the other three conventional sample selection methods. In addition, a comparison of full dataset training with incremental learning training demonstrates that the framework improves the recognition rate by nearly 1%, exhibits better recognition performance, significantly shortens the training time, reduces the cost of model updating and iteration, and is more suitable for practical applications. This study also investigates the use of the incremental learning of action classes, achieving an average recognition rate of 88.6%, which facilitates the supplementation of action types according to the demand, and further improves the application value of the action pattern recognition technology based on sEMG signals.

3-D full-field reconstruction of chemically reacting flow towards high-dimension conditions through machine learning

Monitoring chemically reacting flow is crucial for optimizing and controlling the conversion processes in various chemical engineering scenarios. These processes are often influenced by multiple factors, creating a high-dimensional condition space that imposes intensive computational cost to computational fluid dynamics (CFD) methods. We present a machine learning (ML) framework to reconstruct the 3-D scalar field preserving the accuracy and resolution of CFD modeling. We demonstrate the framework’s effectiveness through a case study on biomass combustion in an industrial-scale grate furnace, which exemplifies the challenges of modeling chemically reacting flow in complex geometric domains influenced non-linearly by multiple parameters. The framework implements three core principles: (1) a Reactor-Structure-Resembled Network (RSRNet) reflecting the operational logics of the reactor, (2) an incremental learning strategy based on random Latin Hypercube Sampling, and (3) the incorporation of self-extracted high-level features from data as soft constraints in the loss function. Ground truth data were calculated through a CFD model, with all cases sampled from a 11-variable condition space. Using only 240 randomly sampled cases (around 1 % total cases in the discretized condition space), RSRNet precisely replicated 1-D and 2-D scalar distributions through incremental training. Further reconstruction of 3-D temperature field only used 40 3-D CFD cases, resulting in the training and testing errors decreased to 7.94 × 10-4 and 1.55 × 10-3, respectively, approaching the level of the 2-D reconstruction, with the average error in temperature field predictions not exceeding 50 K. The generalization ability was validated using a separate test dataset with continuously changing excess air ratio and capacity ratio in wide ranges, including some extreme values, and the model successfully captured complex phenomena under unseen conditions.

Effect of incremental exercise training interventions in a Cardiac Rehabilitation Program following myocardial infarction

Abstract Background Exercise-based cardiac rehabilitation is recommended after myocardial infarction (MI) to improve prognosis, quality of life and control of cardiovascular risk factors. However, the specific impact of different exercise training interventions remains incompletely understood. Purpose To study the health outcomes of three incremental exercise training interventions during a Cardiac Rehabilitation Program (CRP) following MI. Methods We implemented a post-MI Phase 2 CRP in 2022. After initial exercise testing, a 3-month, individualized aerobic and strength ambulatory training program was provided by Cardiologists and Physical Medicine and Rehabilitation physicians (Group 1, n=58). Additionally, a subset of patients received one session of supervised teaching of aerobic and strength training by physiotherapists (Group 2, n=19), and another subset received 8-20 sessions of in-hospital supervised training (Group 3, n=13). Exercise testing was repeated at the end of Phase 2 and peak oxygen consumption (peak VO2) was measured. We analysed quality of life (SF-36), depression (PHQ2), anxiety (GAD2), adherence to Mediterranean diet (PREDIMED) and pharmacological treatment (Morisky-Green), and weekly physical activity (IPAQ). Health outcomes were compared at the beginning and after Phase 2 CRP. Results The cohort comprised 90 patients who completed Phase 2 CRP (mean age 61.2±10.8 years, 84.4% male and 47.8% smokers). Reductions in systolic blood pressure (-10.3±13.8mmHg), LDL cholesterol (-60.7±35.4mg/dL), weight (-2.8±6.6kg, -3±7.8%) and body mass index (-1±2.2) were attained (p<0.05). Smoking abstinence was accomplished in 88.4% of smokers and most patients achieved an LDL<55mg/dL (81.1%). Improvements in quality of life (SF-36 +10.4±19.4 points, p<0.001) and anxiety and depression symptoms (PHQ2 -0.5±1.6 points and GAD2 -0.6±1.6 points, p<0.01) were noted. An increase in peak VO2 (+5.12±5.44ml/kg/min, +23.5±27.4% from baseline, p<0.001) and predicted functional capacity (+13.4±23.8%, p<0.001) was achieved. Patients increased their reported weekly physical activity by 2520.3±2935.4 METS/week (+132% from baseline, p<0.001) and most of them were adherent to Mediterranean diet (91.1%) and pharmacological treatment (95.6%). Patients achieved similar health outcomes regardless of their exercise training group, except for patients in Group 3, who experienced a more pronounced improvement in peak VO2 (+8.1±6.4ml/kg/min, +37.8±34.7% from baseline; p<0.05) and a trend towards higher increase in predicted functional capacity (+21.7±19.2%, p<0.25) and greater weight loss (-5.5±9.4kg, -5.9±9.6%; p<0.2). Conclusions Supervised in-hospital training during a Phase 2 CRP provides superior physical fitness improvements and more pronounced weight loss in MI patients. However, a Phase 2 CRP based on ambulatory training provides similar benefits regarding other health outcomes, such as control of cardiovascular risk factors and quality of life amelioration.Exercise training interventions in a CRP

Multivariate prototype representation for domain-generalized incremental learning

Deep learning models often suffer from catastrophic forgetting when fine-tuned with samples of new classes. This issue becomes even more challenging when there is a domain shift between training and testing data. In this paper, we address the critical yet less explored Domain-Generalized Class-Incremental Learning (DGCIL) task. We propose a DGCIL approach designed to memorize old classes, adapt to new classes, and reliably classify objects from unseen domains. Specifically, our loss formulation maintains classification boundaries while suppressing domain-specific information for each class. Without storing old exemplars, we employ knowledge distillation and estimate the drift of old class prototypes as incremental training progresses. Our prototype representations are based on multivariate Normal distributions, with means and covariances continually adapted to reflect evolving model features, providing effective representations for old classes. We then sample pseudo-features for these old classes from the adapted Normal distributions using Cholesky decomposition. Unlike previous pseudo-feature sampling strategies that rely solely on average mean prototypes, our method captures richer semantic variations. Experiments on several benchmarks demonstrate the superior performance of our method compared to the state of the art.

Efficient incremental training using a novel NMT-SMT hybrid framework for translation of low-resource languages.

The data-hungry statistical machine translation (SMT) and neural machine translation (NMT) models offer state-of-the-art results for languages with abundant data resources. However, extensive research is imperative to make these models perform equally well for low-resource languages. This paper proposes a novel approach to integrate the best features of the NMT and SMT systems for improved translation performance of low-resource English-Tamil language pair. The suboptimal NMT model trained with the small parallel corpus translates the monolingual corpus and selects only the best translations, to retrain itself in the next iteration. The proposed method employs the SMT phrase-pair table to determine the best translations, based on the maximum match between the words of the phrase-pair dictionary and each of the individual translations. This repeating cycle of translation and retraining generates a large quasi-parallel corpus, thus making the NMT model more powerful. SMT-integrated incremental training demonstrates a substantial difference in translation performance as compared to the existing approaches for incremental training. The model is strengthened further by adopting a beam search decoding strategy to produce k best possible translations for each input sentence. Empirical findings prove that the proposed model with BLEU scores of 19.56 and 23.49 outperforms the baseline NMT with scores 11.06 and 17.06 for Eng-to-Tam and Tam-to-Eng translations, respectively. METEOR score evaluation further corroborates these results, proving the supremacy of the proposed model.

Enhancing complex upper-limb motor imagery discrimination through an incremental training strategy

Motor Imagery (MI)-based Brain–Computer Interface (BCI) systems are a great technological advance for the recovery of lost movements in people with severe motor impairments. Different Artificial Intelligence (AI) techniques with supervised methods have been explored for MI task discrimination, especially static movements from left and right hands. Due to different factors affecting MI-based Electroencephalography (EEG) signals related to physical and cognitive conditions, the success rate of BCIs is still low. Currently, there is a need to explore the MI of complex movements associated with Activities of Daily Living (ADLs), which brings challenges to the scientific community. In this work, an incremental training methodology for Artificial Neural Networks (ANNs) is proposed for discrimination of complex MI tasks. MI-Rest and MI-Task related to the imagination of manipulating a drinking cup were discriminated by implementing an Action Observation (AO)-based protocol in a first-person 2D virtual reality. Thirty healthy individuals were recruited to evaluate our complex MI classification approach. The incremental training proposed achieves significantly higher performance (p<0.05) compared to the non-incremental training. Additionally, the proposed method is significantly superior compared to widely applied methods for MI task discrimination: Power Spectrum (PS), Common Spatial Patterns (CSP), Filter Bank Common Spatial Patterns (FBCSP), each using four baseline Machine Learning (ML) methods for classification. The results show an improvement between 5 and 20% according to both Accuracy (ACC) and False Positive Rate (FPR). The proposed methodology obtained promising results, useful for AI-based BCI training, which would allow the development of more robust systems for neurorehabilitation purposes.

Purity Skeleton Dynamic Hypergraph Neural Network

Recently, in the field of Hypergraph Neural Networks (HGNNs), the effectiveness of dynamic hypergraph construction has been validated, which aims to reduce structural noise within the hypergraph through embeddings. However, the existing dynamic construction methods fail to notice the reduction of information contained in the hypergraphs during dynamic updates. This limitation undermines the quality of hypergraphs. Moreover, dynamic hypergraphs are constructed from graphs. Several key nodes play a crucial role in graph, but they are overlooked in hypergraphs. In this paper, we propose a Purity Skeleton Dynamic Hypergraph Neural Network (PS-DHGNN) to address the above issues. Firstly, we leverage purity skeleton method to dynamically construct hypergraphs via the fusion embeddings of features and topology simultaneously. This method effectively reduces structural noise and prevents the loss of information. Secondly, we employ an incremental training strategy, which implements a batch training strategy based on the importance of nodes. The key nodes, as the skeleton of hypergraph, are still highly valued. In addition, we utilize a novel loss function for learning structure information between hypergraph and graph. We conduct extensive experiments on node classification and clustering tasks, which demonstrate that our PS-DHGNN outperforms state-of-the-art methods. Note on real-world traffic flow datasets, PS-DHGNN demonstrates excellent performance, which is highly meaningful in practice.

Data and resource aware incremental ML training in support of pervasive applications

Data and resource aware incremental ML training in support of pervasive applications

Incremental federated learning for traffic flow classification in heterogeneous data scenarios

This paper explores the comparative analysis of federated learning (FL) and centralized learning (CL) models in the context of multi-class traffic flow classification for network applications, a timely study in the context of increasing privacy preservation concerns. Unlike existing literature that often omits detailed class-wise performance evaluation, and consistent data handling and feature selection approaches, our study rectifies these gaps by implementing a feed-forward neural network and assessing FL performance under both independent and identically distributed (IID) and non-independent and identically distributed (non-IID) conditions, with a particular focus on incremental training. In our cross-silo experimental setup involving five clients per round, FL models exhibit notable adaptability. Under IID conditions, the accuracy of the FL model peaked at 96.65%, demonstrating its robustness. Moreover, despite the challenges presented by non-IID environments, our FL models demonstrated significant resilience, adapting incrementally over rounds to optimize performance; in most scenarios, our FL models performed comparably to the idealistic CL model regarding multiple well-established metrics. Through a comprehensive traffic flow classification use case, this work (i) contributes to a better understanding of the capabilities and limitations of FL, offering valuable insights for the real-world deployment of FL, and (ii) provides a novel, large, carefully curated traffic flow dataset for the research community.

Link prediction for hypothesis generation: an active curriculum learning infused temporal graph-based approach

Over the last few years Literature-based Discovery (LBD) has regained popularity as a means to enhance the scientific research process. The resurgent interest has spurred the development of supervised and semi-supervised machine learning models aimed at making previously implicit connections between scientific concepts/entities within often extensive bodies of literature explicit—i.e., suggesting novel scientific hypotheses. In doing so, understanding the temporally evolving interactions between these entities can provide valuable information for predicting the future development of entity relationships. However, existing methods often underutilize the latent information embedded in the temporal aspects of the interaction data. Motivated by applications in the food domain—where we aim to connect nutritional information with health-related benefits—we address the hypothesis-generation problem using a temporal graph-based approach. Given that hypothesis generation involves predicting future (i.e., still to be discovered) entity connections, in our view the ability to capture the dynamic evolution of connections over time is pivotal for a robust model. To address this, we introduce THiGER, a novel batch contrastive temporal node-pair embedding method. THiGER excels in providing a more expressive node-pair encoding by effectively harnessing node-pair relationships. Furthermore, we present THiGER-A, an incremental training approach that incorporates an active curriculum learning strategy to mitigate label bias arising from unobserved connections. By progressively training on increasingly challenging and high-utility samples, our approach significantly enhances the performance of the embedding model. Empirical validation of our proposed method demonstrates its effectiveness on established temporal-graph benchmark datasets, as well as on real-world datasets within the food domain.

Exploration on the change process of human glucocorticoid during incremental exercise load training

Glucocorticoid is a kind of hormone synthesized from glycogen, which has many physiological functions such as improving the body’s anti fatigue, anti stress, and promoting growth. Oxidation medicine is a new medical science with dual characteristics of theory and technology, which is based on biological processes. Glucocorticoids can treat many diseases, but there are certain limitations in this process, especially in diseases that play a vital role. The purpose of this paper is to explore the change process of human glucocorticoid in the process of incremental exercise load training. This paper put forward the algorithm of incremental load training and action recognition in sports training, and analyzed the changes of human glucocorticoid (GC) in the process of incremental load training based on this research. The experimental design includes the selection of research subjects, the determination of sample size, and the setting of experimental and control groups. The load was set to start at 10 a.m., with cycling training from full effort to exhaustion. The initial load was 100 W, and it increased by 20 W every 10 min. The experimental results in this paper showed that for the changes of testosterone (T) during exercise, the salivary T secretion rate immediately after exercise was significantly different from that in the resting state at the recovery period of 30 min. For the change of GC during exercise, the secretory rate of saliva GC was 2.84 mmol/min in the resting state; the secretion rate of saliva GC was 2.46 mmol/min after 20 min of exercise and 2.15 mmol/min after 40 min of exercise; the secretion rate was 1.71 mmol/min immediately after exercise and 2.55 mmol/min after 30 min rest; there was a significant difference in the secretion rate of saliva GC for 5 times. In a word, there was no significant change in saliva T/GC value during exercise and rehabilitation under static state, which was related to strong adaptability of the body and no significant stress response.

SEMG-driven Hand Dynamics Estimation with Incremental Online Learning on a Parallel Ultra-Low-Power Microcontroller.

Surface electromyography (sEMG) is a State-of-the-Art (SoA) sensing modality for non-invasive human-machine interfaces for consumer, industrial, and rehabilitation use cases. The main limitation of the current sEMG-driven control policies is the sEMG's inherent variability, especially cross-session due to sensor repositioning; this limits the generalization of the Machine/Deep Learning (ML/DL) in charge of the signal-to-command mapping. The other hot front on the ML/DL side of sEMG-driven control is the shift from the classification of fixed hand positions to the regression of hand kinematics and dynamics, promising a more versatile and fluid control. We present an incremental online-training strategy for sEMG-based estimation of simultaneous multi-finger forces, using a small Temporal Convolutional Network suitable for embedded learning-on-device. We validate our method on the HYSER dataset, cross-day. Our incremental online training reaches a cross-day Mean Absolute Error (MAE) of (9.58 ± 3.89)% of the Maximum Voluntary Contraction on HYSER's RANDOM dataset of improvised, non-predefined force sequences, which is the most challenging and closest to real scenarios. This MAE is on par with an accuracy-oriented, non-embeddable offline training exploiting more epochs. Further, we demonstrate that our online training approach can be deployed on the GAP9 ultra-low power microcontroller, obtaining a latency of 1.49 ms and an energy draw of just 40.4 uJ per forward-backward-update step. These results show that our solution fits the requirements for accurate and real-time incremental training-on-device.

DEVELOPMENT OF AN ELECTRONIC EXAMINATION PLATFORM USING FACE RECOGNITION METHODS

Online systems face a major challenge in efficiently monitoring participants and students throughout lectures, particularly during exams. Establishing robust approaches and technologies is essential for identifying unjust, unethical, and illegal conduct in educational environments, particularly during academic courses and examinations. This study introduces a novel online proctoring system that employs deep learning to continuously supervise physical locations without requiring a human proctor. The system utilizes face detection and recognition algorithms to implement biometric procedures, including facial recognition. This work presents a novel approach for face recognition training, which involves an incremental training procedure. This eliminates the requirement for an additional step, resulting in reduced computation cost and time. In order to achieve high accuracy, the suggested model evaluated three distinct face detectors: HOG, MTCNN, and Yoloface. The assessment of the suggested model demonstrates that the HOG approach has surpassed the others. The proposed approach yields a significantly high accuracy rate.

The mathematical word problem-solving performance gap between children with and without math difficulties: does working memory mediate and/or moderate treatment effects?

ABSTRACT This study determined the extent to which working memory (WM) played a moderating and/or mediating role in word-problem-solving (WPS) instructional outcomes between children with and without math difficulties (MD). A randomized pretest-posttest control group study investigated the effects of 8-week strategy instruction in one of four treatment conditions on WPS accuracy of third graders with MD (N = 136) when compared to children with (N = 28) and without MD (N = 43). Comparisons were made of three strategy conditions that included overt cues (e.g. underlining key sentences, filling in diagrams), another treatment condition that removed the overt cues (material-only), and two control conditions (children with and without MD). Four important findings emerged. First, posttest WM significantly predicted posttest WPS, computation, and schema accuracy independent of pretest and treatment conditions. Second, posttest WM mediated posttest WPS treatment outcomes when the control conditions included children without MD. Third, strategy conditions that included overt cues (e.g. crossing out irrelevant sentences) decreased WM demands compared to the Materials-Only condition (without overt cues) for children with MD. Finally, incremental attention allocation training within treatment conditions improved posttest WM in children with MD but not posttest WPS. Results indicated that WPS differences between children with and without MD across treatment conditions were mediated by posttest WM performance.

Efficient distributed continual learning for steering experiments in real-time

Deep learning has emerged as a powerful method for extracting valuable information from large volumes of data. However, when new training data arrives continuously (i.e., is not fully available from the beginning), incremental training suffers from catastrophic forgetting (i.e., new patterns are reinforced at the expense of previously acquired knowledge). Training from scratch each time new training data becomes available would result in extremely long training times and massive data accumulation. Rehearsal-based continual learning has shown promise for addressing the catastrophic forgetting challenge, but research to date has not addressed performance and scalability. To fill this gap, we propose an approach based on a distributed rehearsal buffer that efficiently complements data-parallel training on multiple GPUs to achieve high accuracy, short runtime, and scalability. It leverages a set of buffers (local to each GPU) and uses several asynchronous techniques for updating these local buffers in an embarrassingly parallel fashion, all while handling the communication overheads necessary to augment input minibatches using unbiased, global sampling. We further propose a generalization of rehearsal buffers to support both classification and generative learning tasks, as well as more advanced rehearsal strategies (notably Dark Experience Replay, leveraging knowledge distillation). We illustrate this approach with a real-life HPC streaming application from the domain of ptychographic image reconstruction. We run extensive experiments on up to 128 GPUs of the ThetaGPU supercomputer to compare our approach with baselines representative of training-from-scratch (the upper bound in terms of accuracy) and incremental training (the lower bound). Results show that rehearsal-based continual learning achieves a top-5 validation accuracy close to the upper bound, while simultaneously exhibiting a runtime close to the lower bound.

Enhancing graph convolutional networks with progressive granular ball sampling fusion: A novel approach to efficient and accurate GCN training

Graph convolutional network (GCN) has gained considerable attention and has been widely utilized in graph data analytics. However, training large GCNs presents considerable challenges owing to the inherent complexity of graph-structured data. Previous training algorithms frequently struggle with slow convergence speed caused by full-batch gradient descent on entire graphs and reduced model performance due to inappropriate node sampling methods. To address these issues, we propose a novel framework called Progressive Granular Ball Sampling Fusion (PGBSF). PGBSF leverages granular ball sampling to partition the original graph into a collection of subgraphs, thereby enhancing both training efficiency and detail capture. Then, it applies a progressive approach accompanied by a parameter-sharing strategy for incremental GCN model training, which results in robust performance and rapid convergence speed. This simple yet effective strategy considerably enhances classification accuracy and memory efficiency. The experiment results show that our proposed architecture consistently outperforms other baseline models in terms of accuracy across almost all datasets with different label rates. In addition, PGBSF improves GCN performance significantly on large and complex datasets. Moreover, GCN+PGBSF reduces time complexity by training on subgraphs and achieves the fastest convergence speed among all models, with a relatively small variance in loss during training.

Two-dimensional hybrid incremental learning (2DHIL) framework for semantic segmentation of skin tissues

This study aims to enhance the robustness and generalization capability of a deep learning transformer model used for segmenting skin carcinomas and tissues through the introduction of incremental learning. Deep learning AI models demonstrate their claimed performance only for tasks and data types for which they are specifically trained. Their performance is severely challenged for the test cases which are not similar to training data thus questioning their robustness and ability to generalize. Moreover, these models require an enormous amount of annotated data for training to achieve desired performance. The availability of large annotated data, particularly for medical applications, is itself a challenge. Despite efforts to alleviate this limitation through techniques like data augmentation, transfer learning, and few-shot training, the challenge persists. To address this, we propose refining the models incrementally as new classes are discovered and more data becomes available, emulating the human learning process. However, deep learning models face the challenge of catastrophic forgetting during incremental training. Therefore, we introduce a two-dimensional hybrid incremental learning framework for segmenting non-melanoma skin cancers and tissues from histopathology images. Our approach involves progressively adding new classes and introducing data of varying specifications to introduce adaptability in the models. We also employ a combination of loss functions to facilitate new learning and mitigate catastrophic forgetting. Our extended experiments demonstrate significant improvements, with an F1 score reaching 91.78, mIoU of 93.00, and an average accuracy of 95%. These findings highlight the effectiveness of our incremental learning strategy in enhancing the robustness and generalization of deep learning segmentation models while mitigating catastrophic forgetting.

Incremental Load Respiratory Muscle Training Improves Respiratory Muscle Strength and Pulmonary Function in Children with Bronchiectasis.

Participants underwent respiratory muscle training for 24 weeks. The main results were changes in respiratory muscle strength and pulmonary function indices (forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), FEV1/FVC, peak expiratory flow rate (PEF), forced expiratory flow 25-75% (FEF25-75%), and maximal midexpiratory flow 75/25 (MMEF75/25)) before, 12 weeks after, and 24 weeks after the intervention. The secondary outcomes were changes in the exercise load and work rate, exercise work, Leicester Cough Questionnaire (LCQ) scale, and Fatigue Severity Scale (FSS). Compared with before the intervention, after 24 weeks of respiratory muscle training, the maximal inspiratory pressure (MIP) and maximal expiratory pressure (MEP) were significantly enhanced (P < 0.05), while FVC, FEV1, and PEF were significantly increased (P < 0.01). FEF25-75 and MMEF75/25 values showed significant improvement compared to those before training (P < 0.05). The exercise loading, work, and exercise work rate of expiratory muscle training were significantly improved compared to those before intervention (P < 0.05). The LCQ score increased significantly (P < 0.001), and the FSS score decreased significantly (P < 0.001). Incremental load respiratory muscle training effectively improved children's lung function over the long term, improved the strength of their inspiratory and expiratory muscles, and improved their quality of life.

Biomarkers of in vivo platelet activation in thoroughbreds during their first long-term training.

Physical exercise has an activating effect on platelet function that differs between trained and untrained subjects, depending on the type of exercise and training status. In humans, soluble P-selectin (sP-sel) and platelet-derived extracellular vesicles (PEVs) are considered reliable markers of in vivo platelet activation during exercise. In untrained humans, they increase after transient physical exercise, whereas long-term training induces a decrease in their resting levels due to an improved ability to adapt to hemodynamic changes. The aim of this study was to assess whether circulating levels of sP-sel and PEVs may be useful markers to explore in vivo platelet function in never-trained Thoroughbreds during their first 4 months of incremental training. A total of 29 clinically healthy, untrained Thoroughbreds (17 males and 12 females) were enrolled. All horses were trained with the same training schedule (90 days). Blood samples were collected on the day the training program began (T0), 30 days (T30), and 90 days (T90) after its incremental increase to quantify platelet count, sP-sel (horse enzyme-linked immunosorbent assay) and PEVs (flow cytometry). Statistical analysis was performed using RM one-way analysis of variance with the Geisser-Greenhouse correction. Soluble P-selectin tended to increase at T30 compared with T0, while T90 levels returned to baseline values. Significantly higher circulating levels of PEVs CD61+/AnnV+ were observed at T30 and T90 compared to baseline confirming platelet hyperactivity. The detection and quantification of sP-sel and PEVs in equine racehorses during the training period appears to be a promising tool to study exercise-induced primary hemostatic changes and may provide an important marker for exercise selection.

The Effect of Endurance Training with Garlic and Stevia Supplementation on AgRP and α-MSH in Brain Tissue of Obese Wistar Rats

Background: The aim of this study is to investigate the effect of endurance training with garlic and stevia supplementation on AgRP and α-MSH in the brain tissue of obese Wistar rats. Methods: 36 obese male Wistar rats fed with a high-fat diet (24 g fat, 24 g protein, and 41 g carbohydrate/100 g) were divided into six groups, including: 1) sham; 2) stevia supplementation; 3) garlic supplementation; 4) endurance training; 5) stevia supplementation plus endurance training, and 6) garlic supplementation plus endurance training. Moreover, 6 rats were included in the healthy control group to investigate the effects of obesity on the research variable. Training groups performed incremental endurance training for 15-50 minutes at a speed of 15-25 m/min, for eight weeks. Garlic and stevia supplements were added to the diet of the supplementation groups at a dose of 250 mg/kg. Results: Levels of AgRP in stevia supplementation group were significantly higher than endurance training (P=0.019) and endurance training plus stevia supplementation (P=0.018) groups. Levels of α-MSH in the garlic supplementation (P=0.001), endurance training (P=0.002), stevia supplementation plus endurance training (P=0.001), and garlic supplementation plus endurance training (P=0.001) groups were significantly higher than the sham group. In addition, in the endurance training plus stevia supplementation group, the levels were significantly higher than the stevia supplementation group (P=0.002). Conclusion: Endurance training plus garlic supplementation as well as endurance training plus stevia supplementation seem to play a synergistic role in appetite control protein; however, more studies on AgRP changes, following training and garlic and stevia supplementation are needed.