Learning Curve Research Articles

Does the experience of the endoscopic vein harvester matter to the quality of the vein conduit: A critical thematic literature review.

Coronary artery bypass graft surgery remains the golden standard surgical option for multiple vessel disease. Harvesting the long saphenous vein using endoscopic vein harvesting requires advanced surgical skills dexterity, but the lack of a national standardised training programme allows for variance in the learning curve and the quality of the vein during the learning cycle is unknown. A search of bibliographic databases: CINHAL Plus, Embase, Pubmed and the Cochrane register for randomised controlled trials identified 11 articles eligible for review. The themes emerging were learning curve-associated injuries to the long saphenous vein, intimal wall remodelling of the long saphenous vein and incidence of graft patency rates. Harvesting practitioners with less than 100 cases of experience inflict more conduit injuries leading to endothelial remodelling and narrowed vein grafts at the six-month point resulting in lumen loss. Practitioners with more than 100 cases demonstrated reduced learning curve-related injuries on the conduit. Adopting a formalised structured training programme such as the Manchester Endoscopic Learning Tool has shown to reduce endothelial injury to the long saphenous vein minimising early vein graft failure during the learning cycle.

Evaluating the efficacy of a cost-effective, fully three-dimensional-printed vertebra model for endoscopic spine surgery training for neurosurgical residents.

A fused deposition modeling three-dimensional (3D)-printed model of the L4-5 vertebra for lumbar discectomy was designed. The model included separately printed dura mater, spinal cord, ligamentum flavum, intervertebral disc (from thermoplastic polyurethane), and bony structures (from polylactic acid), and the material cost approximately US$ 1 per model. A simple plumbing endoscope was used for visualization. Dura mater injury was assessed by painting two layers on the dura mater, which peeled off with trauma. Endoscopic spine surgery is a subject of high interest in neurosurgery given its minimally invasive nature; however, it has a steep learning curve. This study evaluated the effectiveness of a cost-efficient 3D-printed model when teaching this technique to neurosurgery residents. Only a few studies have investigated the efficacy of such a model. Eight residents with >2 years of training participated. Residents performed the procedure bilaterally and twice at 1-week intervals. From the 32 surgeries, four were excluded because of facet removal (as it widened the surgical corridor), leaving 28 surgeries for analysis. Initial surgeries demonstrated a mean operation time of 21 minutes 18 seconds (standard deviation [SD], 2 minutes 32 seconds), which improved to a mean of 6 minutes 45 seconds (SD, 37 seconds) in the fourth surgery (F(3, 17)=19.18, p <0.0001), demonstrating a significant reduction in surgical time over successive surgeries. The median area with the paint removed decreased, from 161.80 (85.55-217.83) to 95.13 mm2 (12.62-160.54), (F(2.072, Inf)=2.04, p =0.128); however, this was not significant. Resident feedback indicated high satisfaction with the educational value of the model. The developed fully 3D-printed model provides a viable and scalable option for neurosurgical training programs, enhancing the learning experience while maintaining low costs. This model may be an excellent stepping stone for learning lumbar spine endoscopy, acclimating to the two-dimensional view, progressing to cadaver models, and, eventually, independent surgery.

A pragmatic review of cost-effective evaluations of reimbursed medical devices conducted for Japan.

The aim of this study is to examine the characteristics and quality of economic evaluations for reimbursed medical devices in Japan. A pragmatic review of cost-effectiveness evaluations (CEA) and other economic evaluations for medical devices reimbursed in Japan published between January 2010 and December 2023 was conducted using the PubMed (Medline), Japan Medical Abstracts Society (ICHUSHI Web), and the Cost-Effectiveness Analysis Registry of the Tufts Medical Center databases - as well as Google Scholar. Evaluations for devices assessed under the health technology assessment (HTA) system in Japan were also reviewed. Thirty-nine published studies were included with 20 studies (51%) for devices used to treat cardiovascular conditions, 7 studies (18%) for devices used to treat neurological/neurovascular conditions and orthopedic conditions, respectively, and 5 studies (13%) for devices used to treat other types of conditions. The number of published studies for reimbursed medical devices increased from 2017. Nearly 60% of the studies were cost-utility analyses and many were cost-consequence analyses (26%). Although the quality of the studies conducted were good, lack of data was mentioned as a key limitation of nearly all studies - with limited data in general (33%), lack of long-term outcomes data (33%), and lack data for Japan (21%) being key issues. Moreover, lack of cost data was mentioned as a limitation for nearly half (49%) of studies. As this was not a systematic literature review, some relevant studies may have been excluded. Moreover, some databases that are known to cover other journals were not used. Despite concerns about lack of sufficient outcomes data, good quality CEAs have been published for reimbursed devices in Japan. However, lack of data may still be an issue and the impact of the learning curve effect on cost-effectiveness may need to be considered more.

Analysis of the Construction Sequence of a Reaction Wall in a Structural Laboratory Supported by the VDC Methodology

The structures laboratory is a crucial facility for supporting research development related to the construction industry. This kind of laboratory has several components, and one of its key components is the reaction wall. The reaction wall supports the execution of experimental tests on structural elements under specific configurations and loads. Constructing a reaction wall presents challenges due to the highest standards to be met. This research analyzes the construction process of a reaction wall at a private university in Peru. Initially, a schedule was proposed by a contractor. However, the client, knowing the complexity of such construction, asked to review the initial schedule to ensure that it would meet the constructability requirements under the integrated product-organization-process approach given by the virtual design and construction (VDC) methodology. The result of this study pointed out the need to revise the original proposed schedule, extending the construction time to 66% due to the lack of construction processes compatible with the requirements for the reaction wall’s execution. Finally, a survey was conducted to observe the learning curve associated with implementing the VDC methodology.

High-dimensional asymptotics of denoising autoencoders*

Abstract We address the problem of denoising data from a Gaussian mixture using a two-layer non-linear autoencoder with tied weights and a skip connection. We consider the high-dimensional limit where the number of training samples and the input dimension jointly tend to infinity while the number of hidden units remains bounded. We provide closed-form expressions for the denoising mean-squared test error. Building on this result, we quantitatively characterize the advantage of the considered architecture over the autoencoder without the skip connection that relates closely to principal component analysis. We further show that our results accurately capture the learning curves on a range of real data sets.

Enhancing Conceptual Understanding of Electricity and Magnetism through VR Simulations

The study explores the efficacy of Virtual Reality (VR) simulations in enhancing students’ conceptual understanding of electricity and magnetism, using the Labster platform. Traditional teaching methods often need to adequately convey the complexities of these topics, leading to limited student engagement and poor comprehension. The research compares two groups: one receiving traditional instruction and another using VR-based simulations. The study found that students in the experimental group, who used VR simulations, showed a 35% improvement in post-test scores, compared to a 15% improvement in the control group. The VR group also reported higher levels of engagement and found abstract concepts such as electric fields and circuits easier to understand through interactive, immersive learning environments. Teachers noted that VR aided in visualizing difficult concepts, although challenges included a learning curve in technology usage and the need for proper integration into curriculum. The findings suggest that VR, when used as a supplemental tool, has the potential to significantly improve learning outcomes in physics education. However, the study highlights the need for further research into long-term impacts and accessibility issues and strategies to better integrate VR into traditional teaching methods.

Learning curves for deep structured Gaussian feature models*

Abstract In recent years, significant attention in deep learning theory has been devoted to analyzing when models that interpolate their training data can still generalize well to unseen examples. Many insights have been gained from studying models with multiple layers of Gaussian random features, for which one can compute precise generalization asymptotics. However, few works have considered the effect of weight anisotropy; most assume that the random features are generated using independent and identically distributed Gaussian weights, and allow only for structure in the input data. Here, we use the replica trick from statistical physics to derive learning curves for models with many layers of structured Gaussian features. We show that allowing correlations between the rows of the first layer of features can aid generalization, while structure in later layers is generally detrimental. Our results shed light on how weight structure affects generalization in a simple class of solvable models.

Revolutionizing Computer Science Education: Integrating Blockchain for Enhanced Learning and Future Readiness

Abstract: This research investigates the transformative potential of incorporating blockchain technology into computer science education. In light of the rapid evolution of the digital landscape, traditional educational frameworks often fail to meet the demands of the industry, resulting in a significant skills gap among graduates. This paper analyzes how the integration of blockchain can revolutionize computer science curricula by enhancing learning experiences and equipping students to navigate future technological challenges. The study highlights the advantages of blockchain in educational contexts, including increased security and transparency of academic records, streamlined credentialing processes, and the establishment of decentralized learning platforms that promote collaboration and innovation. It presents case studies of institutions that have successfully implemented blockchain, along with strategies for educators to effectively integrate this technology into their pedagogical approaches. Additionally, the research addresses the challenges and limitations associated with blockchain integration, such as the requisite learning curve and infrastructure demands. The findings indicate that the incorporation of blockchain in computer science education can significantly boost student engagement, provide verifiable skill sets, and align academic outcomes more closely with industry requirements. This study contributes to the ongoing discourse on educational innovation and offers a strategic framework for institutions aiming to utilize blockchain technology in preparing students for the future job market.

Intraoperative navigation in craniofacial surgery

Craniofacial surgery requires comprehensive anatomical knowledge of the head and neck to ensure patient safety and surgical precision. Over recent decades, there have been significant advancements in imaging techniques and the development of real-time surgical navigation systems. Intraoperative navigation technology aligns surgical instruments with imaging-derived information on patient anatomy, enabling surgeons to closely follow preoperative plans. This system functions as a radiologic map, improving the accuracy of instrument placement and minimizing surgical complications. The introduction of first-generation navigation systems in the early 1990s revolutionized surgical procedures by enabling real-time tracking of instruments using preoperative imaging. Initially utilized in neurosurgery, intraoperative navigation has since become standard practice in otolaryngology, cranio-maxillofacial surgery, and orthopedics. Since the 2000s, second-generation navigation systems have been developed to meet the growing demand for precision across various surgical specialties. The adoption of these systems in craniofacial surgery has been slower, but their use is increasing, particularly in procedures such as foreign body removal, facial bone fracture reconstruction, tumor resection, and craniofacial reconstruction and implantation. In Korea, insurance coverage for navigation in craniofacial surgery began in 2021, and new medical technologies for orbital wall fracture treatment were approved in August 2022. These technologies have only recently become clinically available, but are expected to play an increasingly important role in craniofacial surgery. Intraoperative navigation enhances operative insight, improves target localization, and increases surgical safety. Although these systems have a steep learning curve and initially prolong surgery, efficiency improves with experience. Calibration issues, registration errors, and soft tissue deformation can introduce inaccuracies. Nonetheless, navigation technology is evolving, and the integration of intraoperative computed tomography data holds promise for further enhancements of surgical accuracy. This paper discusses the various types and applications of navigation employed in craniofacial surgery, highlighting their benefits and limitations.

Subjectivity of visual assessments in FOCUS kinetics and acceptability of first-order fits for regulatory modelling

The degradation half-life (DegT50) of a substance in soil plays an important role in the approval process of a plant protection product and is a sensitive input parameter for regulatory models. It is usually derived through least squares optimizations of mathematical models to measured degradation data according to EU FOCUS degradation kinetics guidance. A strong consensus on degradation parameters provides a solid foundation for parts of the environmental risk assessment. The DegT50 of a substance for regulatory modeling is preferably derived from a single first-order (SFO) model as this is currently the only kinetic model implemented in EU regulatory models of the environmental fate of pesticides. However, kinetic optimisation tools do not always provide a regulatory acceptable SFO fit even though a visual inspection of the data suggests it may be possible. It was therefore hypothesized that more acceptable SFO fits might be achieved by adapting the objective function that is minimized during the optimization.Eight objective functions with varying weightings were tested on 29 laboratory soil degradation datasets. A web-based app was developed to allow experts in environmental safety of plant protection products to visually assess the goodness of fits resulting from different objective functions. The visual assessments and a quantitative metric, newly introduced in the proposed update of the FOCUS guidance, show that the acceptability of SFO fits can be increased, but no single objective function exclusively improves all fits. The assessment reveals that expert judgment is very subjective. Participants tended to change their mind when judging the acceptance of a fit, assumingly caused by a learning curve or a period of calibration.It is concluded that different objective functions could be considered in the kinetic assessment as it can improve the acceptability of SFO fits and hence endpoints for regulatory modeling. This study reveals that various qualitative factors influence the visual judgment of experts when performing a kinetic modeling assessment. The proposed quantitative metric seems to be in alignment with the visual assessment of fits to derive modeling endpoints and a promising step toward less subjective kinetic modeling assessments.

A Realistic Ingeniously Designed Cleft Palate Simulator for Training of Residents: Bridging the Gap Between Theory and Practice.

Cleft palate is a congenital deformity that presents significant challenges in surgical correction. Proper training with an educational model for can greatly enhance the learning curve and improve patient outcomes. This article discusses the development and evaluation of an economical cleft palate model designed to enhance surgical skills and confidence among residents. The model was constructed using dental casts, acrylic resin, foam sheets, double-sided adhesives, infant feeding tubes, and red dye. Tertiary care center. Surgical residents. The assembly time and material costs were recorded. To evaluate its effectiveness, participants' feedbacks were collected and analyzed. The cleft palate model was produced at a cost of 250 INR per unit, with assembly taking approximately 15 minutes. Overall, 32 residents completed the simulation exercise of cleft surgery on the model. Participants gave the model an average score of 4 (±0.60) for realistic anatomical appearance, 3.5 (±1.3) for replicating the tactile nature of human tissue. Anatomical accuracy and ability to manipulate tissue and perform suturing received a rating of 3.4 (±1.1) and 3.7 (±0.81), respectively. The model's value as a training tool scored 4.4 (±0.51), with an overall satisfaction rating of 4.6 (±0.22) among residents. Our cleft palate model offers an economical, practical, and accessible solution for surgical training. This model is a viable alternative to more complex and expensive training tools and potentially enhancing the training of healthcare professionals in cleft palate repair.

From Simulations to Inference: Using Machine Learning to Tune Patient-Specific Finite-Element Models of the Middle Ear Towards Objective Diagnosis.

Computational models, particularly finite-element (FE) models, are essential for interpreting experimental data and predicting system behavior, especially when direct measurements are limited. A major challenge in tuning these models is the large number of parameters involved. Traditional methods, such as one-by-one sensitivity analyses, are time-consuming, subjective, and often return only a single set of parameter values, focusing on reproducing averaged data rather than capturing the full variability of experimental measurements. In this study, we applied simulation-based inference (SBI) using neural posterior estimation (NPE) to tune an FE model of the human middle ear. The training dataset consisted of 10,000 FE simulations of stapes velocity, ear-canal (EC) input impedance, and absorbance, paired with seven FE parameter values randomly sampled within plausible ranges. The neural network learned the association between parameters and simulation outcomes, returning the probability distribution of parameter values that can reproduce experimental data. Our approach successfully identified parameter sets that reproduced three experimental datasets simultaneously. By accounting for experimental noise and variability during training, the method provided a probability distribution of parameters, representing all valid combinations that could fit the data, rather than tuning to averaged values. The network demonstrated robustness to noise and exhibited an efficient learning curve due to the large training dataset. SBI offers an objective alternative to laborious sensitivity analyses, providing probability distributions for each parameter and uncovering interactions between them. This method can be applied to any biological FE model, and we demonstrated its effectiveness using a middle-ear model. Importantly, it holds promise for objective differential diagnosis of conductive hearing loss by providing insight into the mechanical properties of the middle ear.

Immersive virtual reality in forestry work – implementation for chainsaw training

ABSTRACT Growing interest in virtual reality (VR) and its technical development has resulted in several applications for training and teaching purposes. This paper studied an immersive VR platform, hereafter referred to as IVR, developed for training in motor-manual tree delimbing as a learning tool, which is one of the first in forestry work training. The tested platform was the Husqvarna Limberjack, available for free download and publicly accessible. The platform uses an immersive VR to teach users how to delimb stems with a chainsaw. The study included four groups with different previous experience of using immersive VR as the first factor and a chainsaw as the second factor. The experiment included records of the time required to complete the tasks. That data was used in a procedural study in which the average and best times of the groups were utilized. Statistical analysis showed differences between the groups with and without experience in both factors. The learning curves showed a 12.4 to 26.7% improvement for all four groups. The IVR platform turned out to be a useful training aid, where skills can be gained before going to practice in field.

Dynamic Structural Equation Modeling with Cycles

Cyclical phenomena are commonly observed in many areas of repeated measurements, especially with intensive longitudinal data. A typical example is circadian (24-hour) rhythm of physical measures such as blood pressure, heart rate, glucose level, and alertness. This paper focuses on positive affect, which is a common measure in psychological studies and for which circadian rhythm has been observed but not analyzed by modern statistical methods. The paper demonstrates that a large new analysis arsenal is available for analysis of cyclical features in intensive longitudinal data. This can help researchers extract more information from their data to get more valid estimates of coupled processes and to get new theoretical insights into circadian rhythms of mood. To assist in this effort, the analyses are based on general models with a rich set of features while still being accessible without an unduly steep learning curve. Scripts for the Mplus software are available for all the analyses presented.

Innovating Against Odds: A Medical Student's Research Journey in a Resource-Constrained Environment

"You need not know the R in Research to embark on the journey; a strong motivation is enough to set things in motion and let the journey unfold naturally." In this article, I share my journey as a medical student starting research in a Resource-Constrained Environment. I talk about the learning curve, milestones, setbacks, and presenting my research work. I've created a simple guide for fellow students in similar situations to do impactful research. Beyond studies, the shared journey with my research partner formed a lasting bond, showing how teamwork fosters understanding and innovation. I also discuss institutional barriers, especially the toxic research culture not talked about much. I stress the importance of raising awareness about these challenges for medical students and the need for a supportive and ethical research environment. It's crucial to create a space where aspiring researchers can thrive without unethical practices, emphasizing the value of knowledge pursuit over personal ambitions.

Endocrine surgery fellowship is necessary for competent endocrine surgical practice: perspectives from Australia and New Zealand.

Currently pathways to practice in endocrine surgery vary based on location and surgical training programme. International data highlights the impact of surgeon volume on outcomes, and the importance of understanding the learning curve in developing minimum training competencies. This study aims to explore how surgeons obtain competence in endocrine surgery in Australia and New Zealand, and perceptions around competence and scope of practice. A web-based survey was distributed to fellows practicing endocrine surgery. Participants were invited to complete a semi-structured interview to explore key themes around competence. Thematic analysis was performed. Responses from 87 surgeons, with 30% practicing primarily in a regional or rural area, showed 94% emphasized post-fellowship training to be competent in endocrine surgery. Median primary operator procedural volume learning curves were 50 thyroid, 30 parathyroid and 20 laparoscopic adrenalectomy procedures. Semi-structured interviews with 12 participants identified four major themes: (1) learning opportunities during general surgical education and training programmes alone are insufficient for consultant-level competence; (2) the importance of sufficient training to develop clinical decision-making, insight and judgement to appropriately select patients in the management of endocrine disease; (3) expected standards of clinical and technical performance are independent of practice location or context; (4) the importance of multi-disciplinary teams for complex cases including advanced cancers. Practicing endocrine surgeons acknowledge formal fellowship training is required to achieve competence across technical and non-technical domains. The definition of competence and expectations regarding technical outcomes are independent of practice location or context.

A Systematic Literature Review on Augmented Reality in Game-Based Learning for English as a Second Language

This systematic review explores the integration of Augmented Reality (AR) within game-based learning for English as a Second Language (ESL). Based on empirical studies spanning 2020 to 2024 sourced from Web of Science (WoS), Scopus, ERIC, and Google Scholar. The review assesses AR's impact on ESL education. The findings highlight AR's potential to enhance language attainment through mobile applications, fostering learner satisfaction, engagement, and motivation. Additionally, the fusion of game-based learning and AR diversifies learning experiences across subjects, promoting collaboration and communication skills. However, challenges such as technical hurdles, the learning curve for teachers and students, financial constraints, and potential distractions from learning content persist. Despite these obstacles, AR presents promising opportunities for improving ESL learning outcomes. Moreover, it contributes to understanding AR's role in ESL education and underscores the need for further research to address implementation barriers and maximise its educational potential.

Invited Commentary: Transcarotid Artery Revascularization: Short Learning Curve but Low Procedural Volume Still Predicts Higher Morbidity.

Invited Commentary: Transcarotid Artery Revascularization: Short Learning Curve but Low Procedural Volume Still Predicts Higher Morbidity.

Force control for the robot-assisted laparoscopic ultrasound scanning system

Laparoscopic ultrasound scanning, an acritical diagnostic imaging method for liver diseases, necessitates precise contact force control that is typically dependent on skilled physicians. Given its high mobility and controllability, the robot is anticipated to supplant physicians in performing laparoscopic ultrasound scans. Consequently, this study introduces a laparoscopic ultrasound scanning system paired with a learning-based force controller. Compared with traditional methods, our approach offers precise control over the contact force between the ultrasound probe and human organ surfaces during scanning, reducing the learning curve for surgeons in laparoscopic liver ultrasound scanning. Acknowledging the complex nonlinear mechanical model involved when the laparoscopic ultrasound probe contacts human organs and tissues, we propose a learning-based method utilizing a long short-term memory network to estimate the contact force. This method allows for the precise estimation of contact force with human organ tissues by analyzing tendon force. The results indicate that the proposed LSTM neural network can accurately fit the mechanical model of the robotic catheter, with a root mean square error of 3.44 mN. Finally, we conducted a force control experiment for ultrasound scanning on a silicone liver model. The results indicate that the proposed method can achieve relatively stable force control for laparoscopic ultrasound scanning, with a maximum error of 2.41 mN during the transient phase of system control and a maximum absolute error of 1.8 mN in the steady state.

Visionary constructs: revolutionizing students internships through BIM and extended reality synergy

ABSTRACT This study investigates the integration of Building Information Modeling (BIM) and Extended Reality (XR) technologies in construction education, focusing on their application in internship programs. Traditional internships often face limitations such as restricted exposure to complex scenarios and geographical constraints. To address these challenges, this research examines how BIM-XR technologies can bridge the gap between theoretical knowledge and practical application in the Architecture, Engineering, and Construction (AEC) disciplines. A mixed-methods approach was employed, combining qualitative data from interviews with students, educators, and industry professionals alongside quantitative survey data. This comprehensive methodology enabled an in-depth assessment of BIM-XR's impact on student learning outcomes and operational efficiency. The findings revealed that BIM-XR integration significantly enhances educational experiences by providing immersive simulations that help students visualize complex construction processes. This leads to better preparedness for real-world challenges, increased engagement, and more effective internships. Additionally, BIM-XR offers practical benefits for industry practices, such as improved project planning, real-time decision-making, and enhanced stakeholder collaboration. However, the study also highlights challenges, including the steep learning curve, limited access to necessary technologies, and high implementation costs. These limitations point to the need for comprehensive training programs, affordable technology options, and institutional support to ensure broader adoption. Looking ahead, the study suggests that further integration of BIM-XR with emerging technologies like Artificial Intelligence (AI), Machine Learning (ML), and the Internet of Things (IoT) could further enhance educational outcomes and industry practices.