Training Model Research Articles

Development and Validation of a Novel, User-Friendly Simulation Phantom for Ultrasound-Guided Transvaginal Pelvic Abscess Drainage.

The increasing integration of point-of-care ultrasound into clinical practice necessitates effective training tools. One such practice is pelvic abscess drainage using a vaginal probe, where an appropriate training model does not exist. To develop and evaluate an economical, user-friendly phantom for simulating pelvic abscess drainage. The phantom was constructed using readily available materials, including ballistic gel and a simulant for purulent material. The phantom was designed to simulate pelvic abscess drainage through a vaginal probe. Ten independent investigators and 2 authors evaluated the phantom's realism and utility using a visual analog scale (VAS), ranging from "not realistic and useful" (0) to "very realistic and useful" (10) for anatomical fidelity, material suitability, and ease of performing the examination. Three evaluators with previous experience in abscess drainage compared the phantom to real-life scenarios using the same VAS scale. The median time required for each health care professional to perform the procedure was also measured. The phantom was constructed at a cost of approximately US$ 30.00. Evaluators scored the phantom highly for realism and utility, with mean VAS values of 8.75, 9, and 9.2 for anatomical fidelity, material properties, and ease of performing the examination, respectively. Three senior radiologists, each with extensive experience in real-life abscess drainage procedures, compared the phantom to their clinical experiences. They rated the phantom as highly realistic, assigning it a mean VAS score of 9.8. The median time to perform the procedure was 1 minute and 51 seconds. The developed phantom effectively simulates pelvic abscess drainage, providing a realistic and practical training tool. This phantom fills a significant gap in current medical simulation resources, offering a cost-effective solution for training health care professionals in essential ultrasound-guided procedures.

Building Capacity: A Systematic Review of Training in the Diagnosis of Autism for Community-Based Clinicians.

In an effort to reduce the "waitlist crisis," researchers have developed training programs to educate community-based clinicians in best-practice autism diagnostic assessments. This systematic review aims to synthesize the effectiveness and implementation outcomes of such trainings. The following databases were searched from inception until August 2023: PubMed, Web of Science, APA PsycINFO, CINAHL, ERIC, and a select number from Google Scholar. Ten studies were included in the present review because they met the following criteria: development and/or evaluation of a training for practicing community-based clinicians to diagnose autism, published full-text in English, and original research. Risk of bias was assessed through an adapted NIH quality assessment tool. Only seven distinct training programs in autism diagnosis for practicing community-based clinicians were identified. Trainings demonstrated preliminary efficacy in the improvement of clinician knowledge, self-efficacy, practice behavior, and diagnostic accuracy. Many of the trainings had a reported positive impact on the community and were feasible to participate in; however, systems-level factors (e.g., time and reimbursement) remain as barriers to community-based diagnosis. Findings from the present review position clinician training as a promising strategy to increase families' timely access to an autism diagnosis. More research on training models is needed due to both the limited number of trainings and the limited reported effectiveness and implementation outcomes. Future implementation studies are also needed to reduce systems-level barriers and to aid in the determination of what trainings best fit the needs of different contexts.

Adapting security and decentralized knowledge enhancement in federated learning using blockchain technology: literature review

Federated Learning (FL) is a promising form of distributed machine learning that preserves privacy by training models locally without sharing raw data. While FL ensures data privacy through collaborative learning, it faces several critical challenges. These include vulnerabilities to reverse engineering, risks to model architecture privacy, susceptibility to model poisoning attacks, threats to data integrity, and the high costs associated with communication and connectivity. This paper presents a comprehensive review of FL, categorizing data partitioning formats into horizontal federated learning, vertical federated learning, and federated transfer learning. Furthermore, it explores the integration of FL with blockchain, leveraging blockchain’s decentralized nature to enhance FL’s security, reliability, and performance. The study reviews existing FL models, identifying key challenges such as privacy risks, communication overhead, model poisoning vulnerabilities, and ethical dilemmas. It evaluates privacy-preserving mechanisms and security strategies in FL, particularly those enabled by blockchain, such as cryptographic methods, decentralized consensus protocols, and tamper-proof data logging. Additionally, the research analyzes regulatory and ethical considerations for adopting blockchain-based FL solutions. Key findings highlight the effectiveness of blockchain in addressing FL challenges, particularly in mitigating model poisoning, ensuring data integrity, and reducing communication costs. The paper concludes with future directions for integrating blockchain and FL, emphasizing areas such as interoperability, lightweight consensus mechanisms, and regulatory compliance.

Development of a Basic Sepak Sila Training Model in Sepak Takraw for Fifth-Grade Elementary Students

Sepak sila, as the most dominant technique in sepak takraw, requires a training model tailored to player characteristics. Therefore, it is essential to develop a training model that meets these needs. This study aims to describe the feasibility, practicality, and effectiveness of the sepak sila basic technique training model in sepak takraw for fifth-grade elementary school students. The research method used is Research and Development (R&D) with the ADDIE development model. The study's subjects were fifth-grade students from SDN Taipa, with trials involving one teacher and 28 students. The results indicate that the development of the sepak sila basic technique training model in sepak takraw is feasible, with a validity score of 91.27% in the "very valid" category. This training model is also practical, scoring 94.44% in the "very practical" category. Additionally, this training model effectively improves fifth-grade students' sepak sila skills, with a Cohen's D effect size value of 7.196 in the "very large effect" category.

Expert opinions on pediatric EEG training for non-epilepsy specialists in sub-Saharan Africa.

To collate expert consensus on essential curriculum content for non-epilepsy specialists in EEG interpretation and safe post-training practice. A qualitative study on pediatric EEG training curricula needs was designed in collaboration with an adult education specialist. Data were collected via interviews from 15 epilepsy experts with training experience across high- to low-income settings. Thematic analysis was used to identify sub-themes. The experts voted on the key statements in a two-round Delphi to ascertain consensus. Twelve aspects of pediatric EEG training were identified and categorized thematically: relevance; exposure to pediatrics; focus on pediatrics; barriers; resource-limited setting; entry skills; best pedagogy; assessment; critical skills; reinforcement of skills; training model; and recommendations. This study was driven by the inadequate access to training in pediatric EEG for non-epilepsy specialists, which is further exacerbated by the lack of epileptologists and neurophysiologists. The outcomes from the expert consensus opinions promoted consolidation, adaptation, and evolution of existing models that are viable for practice and to be used worldwide. The Delphi consensus demonstrated alignment among regionally located specialists towards the promotion of effective and maintained training for non-epilepsy specialists, as well as highlighting barriers that should be considered and addressed.

Application of TransUnet Deep Learning Model for Automatic Segmentation of Cervical Cancer in Small-Field T2WI Images.

Effective segmentation of cervical cancer tissue from magnetic resonance (MR) images is crucial for automatic detection, staging, and treatment planning of cervical cancer. This study develops an innovative deep learning model to enhance the automatic segmentation of cervical cancer lesions. We obtained 4063 T2WI small-field sagittal, coronal, and oblique axial images from 222 patients with pathologically confirmed cervical cancer. Using this dataset, we employed a convolutional neural network (CNN) along with TransUnet models for segmentation training and evaluation of cervical cancer tissues. In this approach, CNNs are leveraged to extract local information from MR images, whereas Transformers capture long-range dependencies related to shape and structural information, which are critical for precise segmentation. Furthermore, we developed three distinct segmentation models based on coronal, axial, and sagittal T2WI within a small field of view using multidirectional MRI techniques. The dice similarity coefficient (DSC) and mean Hausdorff distance (AHD) were used to assess the performance of the models in terms of segmentation accuracy. The average DSC and AHD values obtained using the TransUnet model were 0.7628 and 0.8687, respectively, surpassing those obtained using the U-Net model by margins of 0.0033 and 0.3479, respectively. The proposed TransUnet segmentation model significantly enhances the accuracy of cervical cancer tissue delineation compared to alternative models, demonstrating superior performance in overall segmentation efficacy. This methodology can improve clinical diagnostic efficiency as an automated image analysis tool tailored for cervical cancer diagnosis.

Empirical study on the feasibility of hybrid-flexible training model for developing teachers’ artificial intelligence competence

Empirical study on the feasibility of hybrid-flexible training model for developing teachers’ artificial intelligence competence

Endoscopic tip-control measured by a snare tip soft coagulation training model correlates with endoscopist profile and therapeutic expertise (The HAM Study)

ObjectiveOperator technical skill is recognised as a critical determinant of surgical outcomes. However, no equivalent recognition for quality of endoscope tip manipulation (tip-control) exists. We aimed to create an ex-vivo...

Virtual Patient Simulations Using Social Robotics Combined With Large Language Models for Clinical Reasoning Training in Medical Education: Mixed Methods Study.

Virtual patients (VPs) are computer-based simulations of clinical scenarios used in health professions education to address various learning outcomes, including clinical reasoning (CR). CR is a crucial skill for health care practitioners, and its inadequacy can compromise patient safety. Recent advancements in large language models (LLMs) and social robots have introduced new possibilities for enhancing VP interactivity and realism. However, their application in VP simulations has been limited, and no studies have investigated the effectiveness of combining LLMs with social robots for CR training. The aim of the study is to explore the potential added value of a social robotic VP platform combined with an LLM compared to a conventional computer-based VP modality for CR training of medical students. A Swedish explorative proof-of-concept study was conducted between May and July 2023, combining quantitative and qualitative methodology. In total, 15 medical students from Karolinska Institutet and an international exchange program completed a VP case in a social robotic platform and a computer-based semilinear platform. Students' self-perceived VP experience focusing on CR training was assessed using a previously developed index, and paired 2-tailed t test was used to compare mean scores (scales from 1 to 5) between the platforms. Moreover, in-depth interviews were conducted with 8 medical students. The social robotic platform was perceived as more authentic (mean 4.5, SD 0.7 vs mean 3.9, SD 0.5; odds ratio [OR] 2.9, 95% CI 0.0-1.0; P=.04) and provided a beneficial overall learning effect (mean 4.4, SD 0.6 versus mean 4.1, SD 0.6; OR 3.7, 95% CI 0.1-0.5; P=.01) compared with the computer-based platform. Qualitative analysis revealed 4 themes, wherein students experienced the social robot as superior to the computer-based platform in training CR, communication, and emotional skills. Limitations related to technical and user-related aspects were identified, and suggestions for improvements included enhanced facial expressions and VP cases simulating multiple personalities. A social robotic platform enhanced by an LLM may provide an authentic and engaging learning experience for medical students in the context of VP simulations for training CR. Beyond its limitations, several aspects of potential improvement were identified for the social robotic platform, lending promise for this technology as a means toward the attainment of learning outcomes within medical education curricula.

Development and validation of a deep learning algorithm for prediction of pediatric recurrent intussusception in ultrasound images and radiographs

PurposesTo develop a predictive model for recurrent intussusception based on abdominal ultrasound (US) images and abdominal radiographs.MethodsA total of 3665 cases of intussusception were retrospectively collected from January 2017 to December 2022. The cohort was randomly assigned to training and validation sets at a 6:4 ratio. Two types of images were processed: abdominal grayscale US images and abdominal radiographs. These images served as inputs for the deep learning algorithm and were individually processed by five detection models for training, with each model predicting its respective categories and probabilities. The optimal models were selected individually for decision fusion to obtain the final predicted categories and their probabilities.ResultsWith US, the VGG11 model showed the best performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.669 (95% CI: 0.635–0.702). In contrast, with radiographs, the ResNet18 model excelled with an AUC of 0.809 (95% CI: 0.776–0.841). We then employed two fusion methods. In the averaging fusion method, the two models were combined to reach a diagnostic decision. Specifically, a soft voting scheme was used to average the probabilities predicted by each model, resulting in an AUC of 0.877 (95% CI: 0.846–0.908). In the stacking fusion method, a meta-model was built based on the predictions of the two optimal models. This approach notably enhanced the overall predictive performance, with LightGBM emerging as the top performer, achieving an AUC of 0.897 (95% CI: 0.869–0.925). Both fusion methods demonstrated excellent performance.ConclusionsDeep learning algorithms developed using multimodal medical imaging may help predict recurrent intussusception.Clinical trial numberNot applicable.

Development Of The 2022 Pencak Silat Pull Training Model Using Resistance Bands In The Pre-Competition Phase

Changes to the pencak silat rules in 2022 have had a significant impact on techniques in sparring category pencak silat matches, namely the fall technique with a pull. However, to fulfill the demands of the task when playing, the need for adjustments to techniques during training is the key to the success of techniques when playing. Therefore, researchers are interested in developing a variation of the fall technique model with a pull. The purpose of this research is to develop a variation of the pull training model using resistance bands to support techniques in addressing the 2022 pencak silat regulations. This research uses the Research and Development (R&D) method. The results of this study are 11 variations of the pull training model with resistance bands. The results of the martial arts expert validation were 90% declared excellent, meaning that it is feasible to be applied in pre-competition phase training activities. The results of the respondent test in the small group trial showed 90% in the excellent category and the large group trial showed 91% categorized as excellent. The results of the product validity test were declared feasible to use.

An Enhanced Combinatorial Contextual Neural Bandit Approach for Client Selection in Federated Learning

In the dynamic realm of machine learning (ML), federated learning (FL) emerges as a pivotal method for training models on decentralized devices without the need for central data aggregation. This technique confronts the challenges of data heterogeneity, which disrupts the independent and identically distributed (IID) assumptions, adversely affecting the accuracy of the overall model. To tackle this issue, we introduce the federated non-performing-node-resilient neural selector (FNNS), an advanced client selection algorithm grounded in a combinatorial contextual neural bandit framework. This algorithm enhances the extraction of contextual data by assessing each local client using a universally standardized dataset, thereby providing a deeper, context-specific insight suitable for federated environments. In addition, we introduce selection robustness score (SRS), a novel metric designed to quantify the efficacy of client selection in the presence of non-performing-nodes (NPN) conditions. Using this metric, we demonstrate FANS’s effectiveness in enhancing the FL process. Empirical evaluations across diverse settings reveal our method’s superiority over current state-of-theart solutions, with significant improvements in both SRS and global model accuracy.

Efficient Training of Neural Network Potentials for Chemical and Enzymatic Reactions by Continual Learning.

Machine learning (ML) methods have emerged as an efficient surrogate for high-level electronic structure theory, offering precision and computational efficiency. However, the vast conformational and chemical space remains challenging when constructing a general force field. Training data sets typically cover only a limited region of this space, resulting in poor extrapolation performance. Traditional strategies must address this problem by training models from scratch using old and new data sets. In addition, model transferability is crucial for general force field construction. Existing ML force fields, designed for closed systems with no external environmental potential, exhibit limited transferability to complex condensed phase systems such as enzymatic reactions, resulting in inferior performance and high memory costs. Our ML/MM model, based on the Taylor expansion of the electrostatic operator, showed high transferability between reactions in several simple solvents. This work extends the strategy to enzymatic reactions to explore the transferability between more complex heterogeneous environments. In addition, we also apply continual learning strategies based on memory data sets to enable autonomous and on-the-fly training on a continuous stream of new data. By combining these two methods, we can efficiently construct a force field that can be applied to chemical reactions in various environmental media.

High Simulation Training System for Spinal Full-Endoscopic Surgery, Based on the Combination of VR/AR and Magneto-Optical Navigation Technology.

Spinal full-endoscopic surgery is a challenging technique with a steep learning curve, limited by inadequate training models and the shortcomings of cadaver-based training. To address this, we propose a high-simulation training system using VR/AR and magneto-optical navigation technology to enhance skill development and reduce the learning curve. A new simulation training system for spinal full-endoscopic surgery was established, which was conducted by using the data of Chinese Digital Human with medical image parameters for the three-dimensional (3D) reconstruction, as well as integrating the technical advantages of VR/AR, 3D printing, and magneto-optical navigation technology. Based on the original dataset of Chinese digital humans and clinical medical imaging processing, the data model was obtained and then the 3D printing engineering model was created. The simulation perspective filming technology, joint process shaping and cutting technology, and tube placement technology have been constructed, based on the combination of VR/AR under optical navigation. Based on electromagnetic tracking, a microscopic anatomical simulation using preorder optical navigation has been designed. Finally, a physical simulation model based on the clinical reality of flocculent behavior was constructed. As a simulator for spinal endoscopic surgery training, it was tested during an advanced endoscopic training course. Moreover, 17 out of 20 novices (85%) met the surgical standards by the end of the final simulation training session. We present a high simulation training system based on the combination of VR/AR and magneto-optical navigation for spinal full-endoscopic surgery. The model may be used by surgeons starting with spinal endoscopy and should be considered a comparable and sufficiently realistic tool to train key operation steps to reduce the learning curve.

Microanastomosis in large discrepancies: a novel training model

Microanastomosis in large discrepancies: a novel training model

Remaining useful life prediction of rolling bearing based on anomaly correction

ABSTRACT The remaining useful life (RUL) prediction based on deep learning obtains certain prediction results by training model with deterministic parameters. However, this method ignores the error range of parameter estimation, resulting in the inability to fully characterise the uncertainty. Therefore, a RUL prediction method based on anomaly correction is proposed. Firstly, to address the problem of RMS deviating from the degradation trajectory caused by external random factors, this paper proposes an adaptive time window anomaly correction strategy with sliding windows are used to construct degenerate slope. Secondly, to solve the problem of prediction result errors caused by changing random factors, this paper considers heteroscedasticity in prognosis and constructs a probability density function library to dynamically match the probability distribution. Then, aiming at the mean value as the prediction result may lead to overconfidence, a probability-based weighted enhanced method is proposed to improve the prediction accuracy. Finally, two datasets are used to verify the effectiveness and superiority of the proposed method.

Effects of Polarized Training vs. Other Training Intensity Distribution Models on Physiological Variables and Endurance Performance in Different-Level Endurance Athletes: A Scoping Review.

Rivera-Köfler, T, Varela-Sanz, A, Padrón-Cabo, A, Giráldez-García, MA, and Muñoz-Pérez, I. Effects of polarized training vs. other training intensity distribution models on physiological variables and endurance performance in different-level endurance athletes: a scoping review. J Strength Cond Res 39(3): 373-385, 2025-This scoping review aimed to analyze the long-term effects of polarized training (POL) on key endurance physiological- and performance-related variables and to systematically compare them with other training intensity distribution (TID) models in endurance athletes of different performance levels. Four TID models were analyzed: POL, pyramidal (PYR), threshold (THR), and block (BT) training models. The literature search was performed using PubMed, SportDiscus, Scopus, and Web of Science databases. Studies were selected if they met the following criteria: compared POL with any other TID model, included healthy endurance athletes, men, and/or women; reported enough information regarding the volume distribution in the different training intensity zones (i.e., zone 1, zone 2, and zone 3), assessed physiological (i.e., maximum/peak oxygen uptake, speed or power at aerobic and anaerobic thresholds, economy of movement), and performance in competition or time-trial variables. Of the 620 studies identified, 15 met the eligibility criteria and were included in this review. According to scientific evidence, POL and PYR models reported greater maximum oxygen uptake enhancements. Both POL and PYR models improved the speed or power associated with the aerobic threshold. By contrast, all TID models effectively improved the speed or power associated with the anaerobic threshold. Further research is needed to establish the effects of TID models on the economy of movement. All TID models were effective in enhancing competitive endurance performance, but testing protocols were quite heterogeneous. The POL and PYR models seem to be more effective in elite and world-class athletes, whereas there were no differences between TID models in lower-level athletes.

Implementation of a cascade training model to enhance emergency care capacity of healthcare workers during the COVID-19 outbreak in Uganda.

Implementation of a cascade training model to enhance emergency care capacity of healthcare workers during the COVID-19 outbreak in Uganda.

Three Types of Side-to-Side Microvascular Anastomosis Training Models Using Rat Abdominal Vessels.

Side-to-side microvascular anastomosis is the most difficult type of anastomosis. The best way to master microvascular anastomosis technique is deliberate practice in the microsurgical laboratory. Three types of side-to-side microvascular anastomosis using the rat abdominal vessels were presented. First, we present step-by-step technical details of common iliac artery (CIA)-CIA side-to-side microvascular anastomosis between 2 perfectly matched common iliac arteries via the in situ intraluminal suturing technique. Then, we present 2 arteriovenous side-to-side anastomosis training models using the same suturing technique: one with the CIA-common iliac vein (CIV) (CIA-CIV anastomosis) and the other with the abdominal aorta (AA) and inferior vena cava (IVC) (AA-IVC anastomosis). Diameters of CIA, CIV, AA, and IVC; the length of arteriotomy or venotomy; and the suturing time were recorded. The patency rates were evaluated immediately after the anastomosis was completed and 30 minutes later. In CIA-CIA side-to-side anastomosis, the bilateral CIAs were perfectly matched in terms of thickness, texture, and consistency. The vascular walls of the CIV and IVC were very thin, and they became transparent and collapsed after being cut open. A small-diameter thick-walled artery was anastomosed to a large-diameter, very thin, and fragile vein in the CIA-CIV and AA-IVC arteriovenous anastomoses. Three types of side-to-side anastomosis using the rat abdominal vessels were successfully performed; 100% patency rates were achieved immediately and 30 minutes postoperatively. Three types of side-to-side microvascular anastomosis training models using abdominal vessels could be used to mimic different microvascular anastomotic situations.

Finding potentially erroneous entries in METLIN SMRT.

Finding potentially erroneous entries in METLIN SMRT.