Value Of Training Research Articles

Advancing Voice Anti-Spoofing Systems: Self-Supervised Learning and Indonesian Dataset Integration for Enhanced Generalization

This study examines how self-supervised learning and a novel Indonesian language dataset enhance anti-spoofing systems. Results show improved model performance, with a lower Equal Error Rate (EER) during training, indicating effective learning from diverse audio samples. Using weighted cross-entropy analysis highlights the model's robustness in minimizing training errors. Comparisons with baseline models using English data reveal the proposed approach's superiority, achieving a significantly lower EER due to the incorporation of language-specific data. The unique phonetic features of Indonesian languages provide valuable training material, boosting the system's defence against spoofing attacks. The dataset improves generalization across dialects and recording conditions by including diverse speech samples. This integration enhances the anti-spoofing systems' adaptability, which is vital for real-world applications where recording variability affects performance. The experimental setup used a balanced dataset of genuine and spoofed utterances from male and female speakers, ensuring high-quality input. The training configuration splits the dataset into training, development, and testing sets on a high-performance computing setup. Results showed the proposed model achieved an EER of 0.33, compared to 7.65 for the traditional sinc-layer model and 0.82 for the wav2vec 2.0 model with English data. Overall, this research advances anti-spoofing solutions and emphasizes the need for diverse datasets and advanced learning approaches to improve automatic speaker verification systems in practical applications. The incorporation of the Indonesian dataset is vital for addressing linguistic diversity challenges in biometric security, paving the way for future advancements in this area.

Meta-Analysis of the Effects of Core Stability Training on 50-Meter Freestyle Performance in Men and Women

Introduction. Core stability training (CST) has gained increasing relevance in sports due to its potential to enhance athletic performance and reduce injury risk, particularly in swimming. The 50-meter freestyle is a key performance test in swimming, and understanding the impact of CST on this specific event is essential for optimizing training programs. Objective. This meta-analysis aimed to evaluate the effects of CST on the performance of male and female swimmers in the 50-meter freestyle. Methods. A systematic search was conducted in PubMed, Bireme, Scopus, and Web of Science (WOS), adhering to the Cochrane Handbook guidelines. Risk of bias was evaluated using the ROB2 scale, while the quality of the studies was assessed with the SIGN and CONSORT checklists. Data were analyzed using a fixed-effects meta-analysis in RevMan-Web, and heterogeneity was assessed using the I2 and X2 tests. Results. Out of 2,323 records identified, 7 studies met the inclusion criteria. The meta-analysis revealed that CST significantly improved 50-meter freestyle performance, with a time reduction of -1.06 seconds (95% CI = -1.52, -0.60) in male swimmers and -3.28 seconds (95% CI = -4.57, -1.99) in female swimmers. Conclusion. CST was found to be effective in enhancing performance in the 50-meter freestyle, particularly in female swimmers. These findings support the use of CST as a valuable training strategy for sports scientists and coaches aiming to improve swimming performance.

Human placenta in vascular neurosurgery training: a comprehensive review of laboratory models, possible exercises, and assessments for skills improvement.

Simulation is increasingly recognized as an important tool to enhance neurosurgical education. In this field, especially in neurovascular surgery, human placenta (hPl) is being recognized as a valuable training model because of its abundant availability, ethical acceptance, and analogous vasculature with other vessels of the human body. Nevertheless, although a rising body of literature is witnessing a renovated interest toward this "old" model, to date no comprehensive reviews on the topics are available. Main objective of this paper is to comprehensively review hPl use in neurovascular surgery training, focusing on model preparation, type of possible exercises along with assessment tools which can be used to check for trainees learning. A systematic review of the pertinent literature was performed, following PRISMA guidelines, searching the following key words in "title/abstract" fields in only English-written works: "placenta and neurosurgery." We focused on preclinical works concerning use of hPl models in neurovascular surgery training. We included studies published in the last 40 years, till December 31st, 2023. A total of 950 scientific publications were initially screened, and 23 papers met the inclusion and exclusion criteria. In summary, the chorionic surface of hPl may be used as a neurovascular gym where different exercises can be performed. Placenta models can be produced with or without the use of intravascular dyes and or/pumping systems to resemble a beating-heart vascular system. The three main neurovascular surgical fields where hPl is being considered to be highly simulative are: 1) Sylvian fissure dissection; 2) aneurysm dissection and clipping; 3) bypass training. Considering the simulative purpose, face, content and construct validify of hPl model have been tested and verified by many authors, although still few data are available regarding transfer/predictive validity. This systematic review highlighted the easiness of preparation and availability along with the strong didactical meaning that the use of hPl-based models may have in the field of neurovascular surgery training. In view of current progressive lowering of surgical vascular cases and increasing technical demand related to their management, further studies aimed to assess if this simulator could be able to produce a learning effect and improve performance with continued use are strongly needed.

Ophthalmic Surgeries at a Nigerian Teaching Hospital from the Perspectives of Undergraduate Medical Education

BackgroundAn appraisal of the profile of ophthalmic surgeries performed at Ladoke Akintola University of Technology Teaching Hospital (LAUTECHTH), Ogbomosho, Nigeria over a seven-year period with a view to harnessing their potentials for maximising ophthalmic experiences of medical students. MethodsA retrospective review of ophthalmic surgery records over a seven-year period (2013 – 2019) at LAUTECHTH. ResultsThere were 386 eye surgeries, mean ophthalmic surgeries: 4 - 5/month or 55/year. Most (329, 85.2%) surgeries were performed on individuals aged at least 45 (range 10 - 95) years with slight male dominance (1.1:1). The most common indication for surgery was cataract (307, 79.6%). Most surgeries were elective (362, 93.8%) and performed under local anaesthesia (367, 95.1%). Despite the highest number of surgeries being performed in the months of April and September; and lowest during the months of May and October; there was fair distribution across and over the years. Small incision cataract surgery with or without intraocular lens implant (307, 79.6%) was the most common surgery followed by pterygium excision with adjuvant therapy (37, 9.6%). ConclusionThe study showed a broad spectrum of eye surgeries, including cataract and pterygium, performed at the teaching hospital, encompassing common conditions that lead to visual impairment or blindness. This exposure provides valuable training opportunities for undergraduate medical students to become physicians especially those aspiring to become ophthalmologists. However, there is a need to increase the overall volume of eye surgeries to enhance the learning experience and potentially influence career choices.

Development and Validation of a Low-Cost, High-Fidelity Simulation Model for Robotic Internal Mammary Artery Harvest Using the da Vinci Xi Robot.

We created and validated a low-cost simulation model for robotic internal mammary artery (IMA) takedown. The simulation model utilized a calf fetus thorax cavity stented open internally and secured to a table. The simulation model was validated at a 2-day robotic cardiac surgery workshop. Each participant harvested one IMA using the da Vinci Xi robot (Intuitive Surgical, Sunnyvale, CA, USA). We compared participant self-reported confidence at robotic IMA harvest before and after using the simulator. Our novel thorax-securing strategy resulted in a stable structure and allowed access to both IMAs from the same 3 ports. The cost to set up the first simulation model was $176 and $133 for every subsequent model. Fifty participants used the simulation model: 42 cardiothoracic surgery attendings and 8 fellows or residents. The feedback form response rate was 78% (n = 39). On the Likert scale, participants rated realism of the calf model to simulate robotic IMA harvesting (0 = not realistic, 10 = highly realistic) with a median of 8 out of 10 (interquartile range [IQR] 7 to 9). Participant confidence (0 = not at all confident, 10 = very confident) in robotic IMA harvesting before and after using the simulator increased (P = 0.001) from a median of 5 (IQR 1 to 7) to 9 (IQR 7 to 10). This robotic IMA harvest simulation model is affordable, realistic, and improved participant confidence in robotic IMA harvest. It may provide a valuable training tool for surgeons learning robotic coronary bypass surgery and allows for training frequency necessary to pass basic learning curves.

More women than ever are entering MD-PhD programs. What lies ahead for them?

The earliest MD-PhD programs were small and enrolled mostly men. Here, we show that since 2014 there has been a steady increase in the number of women in MD-PhD programs, the number of women reaching parity with men in 2023. This change was due to an increase in female applicants, a decrease in male applicants, and an increase in the acceptance rate for women, which had previously been lower than that for men. Data from the National MD-PhD Program Outcomes Study show that training duration has been similar for men and women, as have most choices of medical specialties and workplaces. However, women were less likely to have full-time faculty appointments, fewer had NIH grants, and those in the most recent graduation cohort at the time of the survey reported spending less time on research than men. Previously cited reasons for these differences include disproportionate childcare responsibilities, a paucity of role models, insufficient recognition, and gender bias. Institutions can and should address these obstacles, but training programs can help by preparing their graduates to succeed despite the systemic obstacles. The alternative is a persistent gender gap in the physician-scientist workforce, lost opportunities to benefit from diverse perspectives, and a diminished impact of valuable training resources.

Effects of Bulgarian bag training on bio-motor fitness, physiological, haematological, and performance in young adult male Kabaddi players

Kabaddi demands high bio-motor fitness and physiological robustness in players. The Bulgarian Bag, known for its dynamic training potential, offers multi-planar movements to enhance strength, endurance, and coordination. However, its impact on Kabaddi players still needs to be explored. To assess the effects of Bulgarian Bag training (BBT) on bio-motor fitness, physiological, hematological, and performance measures in young adult Kabaddi players. A total of 30 participants were randomly assigned to either the BBT Group (BBTG) (n = 15; age = 20.73 ± 1.83 years; weight = 72.56 ± 7.27 kg; height = 175.20 ± 6.03 cm) or the active Control Group (CG) (n = 15; age = 20.80 ± 1.69 years; weight = 71.86 ± 4.75 kg; height = 171.46 ± 6.06 cm). The training intervention lasted 12 weeks and was conducted during the in-season period. All outcome parameters tests were conducted before and after the intervention, including agility (AG), explosive strength (ES), shoulder strength (SS), muscular strength (MS), resting heart rate (RHR), VO2max, red blood cells (RBC), white blood cells (WBC), hemoglobin (HG), and playing ability (PA). The ANCOVA analysis revealed significant between-group differences in all physical fitness measures at post-test. Specifically, the BBTG showed significant improvements in AG, ES, SS, MS, RHR, VO2Max, RBC, WBC, HG, and PA (d = 0.75 to 3.28; Δ0.50% to 16.37%). Moreover, the BBTG group significantly reduced resting heart rate (d = 0.75; Δ-3.31%). In contrast, no significant changes were reported in the Control Group (CG) between pre- and post-tests (d = 0.00 to 0.78; Δ-0.08% to 4.59%). Integrating BBTG into the training program of young adult male Kabaddi players yielded beneficial effects across multiple parameters of bio-motor fitness, physiological, hematological, and playing ability. The findings reveal the potential of BBTG as a valuable training method for enhancing Kabaddi performance and overall athletic conditioning. Keywords: Kabaddi Player, Bulgarian bag training, bio motor, physiological, haematological

Game at School? Italian Teachers' Perceptions of the Introduction of Games in the Classroom

The growing awareness that games can be a very valuable training support, both as a resource to engage and motivate and to promote acquisition of disciplinary knowledge and skills, has not been accompanied by an effective diffusion of games in Italian schools. From primary school to university, an ambivalent behaviour is manifested: a widespread appreciation of games as a learning environment is contrasted with a limited application, often reduced to the use of games or online software of a behavioural matrix. An analysis of research on the use and diffusion of games in Italian schools and universities highlights the lack of recent studies on the basic didactic-pedagogical knowledge, level of acceptance and confidence of teachers regarding the introduction of games in the classroom. The goal of this contribution is to provide an updated understanding of Italian teachers' competencies related to the formative use of games. To this aim, we developed a survey adapting to the Italian context one of the most established competence frameworks for game-based learning (GBL): the Technological Pedagogical Content Knowledge-Games (TPACK-G), a variant of TPACK specifically modified for the analysis of games. This new tool extends to non-digital games and integrates the sections on aspects related to game-based instructional design, game scenarios, and the role of the teacher in the instructional activity. This research will not only provide an updated look at how the prevalence of play in the classroom is changing, beyond simple manifestations of interest and trends, but also reveal how the approach to games is changing in the design of educational activities and what is the degree of awareness of the game pedagogical-educational skills necessary for the introduction of games in education. This, in turn, is instrumental for the development of new and more focused teacher training programs to promote the use of games in classrooms.

Creation and evaluation of a three-dimensional-printed synthetic vas deferens for microsurgical training.

Individuals choose to undergo vasectomy reversal for a variety of reasons, including remarriage or due to the death of a child. To be proficient in this procedure, the surgeons need to be high volume and the trainees require a safe environment to develop their microsurgical skills. To overcome this challenge, we used three-dimensional (3D) printing to create a synthetic model of the vas deferens with properties similar to the human vas deferens. We distributed this model to experienced microsurgeons for evaluation. The vas deferens model was created using thermoplastic polyurethane filament. The filament was then infused with a foaming agent to allow for temperature-dependent tuning of the material's stiffness. The model's outer diameter was 1 mm and the inner lumen was 0.5 mm. Fellowship-trained male reproductive urologists were recruited from the Society for the Study of Male Reproduction website. They used our model and judged it on several factors by completing a 13-question survey. We received completed evaluations from five microsurgeons. Eighty percent of the surgeons were able to complete a full anastomosis on the model using 9-0 and 10-0 sutures. The majority of the completed anastomoses were performed using the one-layer technique. The average responses for the model's usefulness as a practice tool, a training tool, and overall assessment ranged from 72 to 79 out of 100. Comments for the improvement included the need for a more flexible and softer model. We created a 3D-printed synthetic vas deferens that serves as a valuable training and practice tool.

A Low-Cost, Low-Tech Virtual Mass Casualty Training Simulation for Undergraduate Medical Education.

Mass casualty incident (MCI) training effectively increases trainees' knowledge and confidence when implemented in a live, in-person setting. In-person MCI training is resource-intensive, but virtual MCI training models are an alternative with similar effectiveness at a lesser cost. However, most of these validated virtual options are based on high-tech virtual reality (VR) programs. We designed and implemented a low-tech, low-cost virtual MCI training model for third-year medical students, using Google JamboardTM (Google, Mountain View, CA) and ZoomTM(Zoom Video Communications, Inc., San Jose, CA) as the primary technological platforms. Learners were instructed on the adult simple triage and rapid treatment (START) and the pediatric JumpSTART triage algorithms over ZoomTM. In small groups, students used a gameboard on Google JamboardTM to simulate a scene at an MCI where they were tasked with triaging 25 patients in 30 minutes, followed by a debriefing session. Students were surveyed on their perceived understanding of the triage algorithms and confidence before and after the event using a 5-point Likert scale (poor, fair, good, very good, and excellent). Pre- and post-event scores were compared using paired, two-sample, and two-tailed t-tests. We considered a p value of <0.01 significant to correct for multiplicity using the Bonferroni method. Learners reported an increased understanding of the triage algorithms (adult and pediatric), scene setup, andunderstanding of emergency medical servicetraining/transportation, as well as increased confidence in participating in an MCI (all p < 0.001). Virtual MCI training can be used as an alternative or supplement to in-person MCI training. Low-tech virtual MCI training models can increase the accessibility of these valuable training activities without sacrificing the quality of learning. Areas for further investigation include low-tech virtual MCI training models' ability to effectively recreate situational and environmental distractions and other challenges better simulated by in-person and high-tech VR training events.

Evaluating User Experience and Data Quality in Gamified Data Collection for Appearance-Based Gaze Estimation

Appearance-based gaze estimation, which uses only a regular camera to estimate human gaze, is important in various application fields. While the technique faces data bias issues, data collection protocol is often demanding, and collecting data from a wide range of participants is difficult. It is an important challenge to design opportunities that allow a diverse range of people to participate while ensuring the quality of the training data. To tackle this challenge, we introduce a novel gamified approach for collecting training data. In this game, two players communicate words via eye gaze through a transparent letter board. Images captured during gameplay serve as valuable training data for gaze estimation models. The game is designed as a physical installation that involves communication between players, and it is expected to attract the interest of diverse participants. We assess the game’s significance on data quality and user experience through a comparative user study.

Memphis NeuroSTART Program: Promoting Student Success and Increasing the Diversity of Applicants to Neuroscience Graduate Programs.

With grant support from the Research Experience for Undergraduates (REU) program funded by the National Science Foundation (NSF) and the Awards to Stimulate and Support Undergraduate Research Experiences (ASSURE) program funded by the Department of Defense (DoD) Air Force Office of Scientific Research (AFOSR), we established a program intended to increase the number of underrepresented racial and ethnic minority (URM) and first-generation undergraduate students successfully applying to neuroscience and other STEM-related graduate programs. The Neuroscience Techniques and Research Training (NeuroSTART) Program aimed to increase the number of undergraduate students from the Memphis area involved in behavioral neuroscience research. In this two-semester program, students completed an empirical research project in a neuroscience lab, received individual mentoring from neuroscience faculty, became part of a STEM network, presented at research conferences, and attended specialized professional development seminars. In two cohorts of 15 students, 4 are PhD students in neuroscience-related programs or in medical school (27%), 4 are employed in neuroscience-related research facilities (27%), 3 are employed as clinical assistants (20%), and 1 is employed in the IT field (7%). The remaining three recently graduated and are planning a gap year prior to applying for admission to graduate/medical school. The Memphis NeuroSTART program has provided valuable training to participants, making them competitive applicants for jobs in the health sciences and for admittance into graduate neuroscience programs. By providing this training to first-generation and URM students, the broader impact of this program was an increase in the diversity of the health sciences workforce, particularly those specializing in neuroscience-related research and treatment.

A Smart Ski Pole for Skiing Pattern Recognition and Quantification Application.

In cross-country skiing, ski poles play a crucial role in technique, propulsion, and overall performance. The kinematic parameters of ski poles can provide valuable information about the skier's technique, which is of great significance for coaches and athletes seeking to improve their skiing performance. In this work, a new smart ski pole is proposed, which combines the uniaxial load cell and the inertial measurement unit (IMU), aiming to provide comprehensive data measurement functions more easily and to play an auxiliary role in training. The ski pole can collect data directly related to skiing technical actions, such as the skier's pole force, pole angle, inertia data, etc., and the system's design, based on wireless transmission, makes the system more convenient to provide comprehensive data acquisition functions, in order to achieve a more simple and efficient use experience. In this experiment, the characteristic data obtained from the ski poles during the Double Poling of three skiers were extracted and the sample t-test was conducted. The results showed that the three skiers had significant differences in pole force, pole angle, and pole time. Spearman correlation analysis was used to analyze the sports data of the people with good performance, and the results showed that the pole force and speed (r = 0.71) and pole support angle (r = 0.76) were significantly correlated. In addition, this study adopted the commonly used inertial sensor data for action recognition, combined with the load cell data as the input of the ski technical action recognition algorithm, and the recognition accuracy of five kinds of cross-country skiing technical actions (Diagonal Stride (DS), Double Poling (DP), Kick Double Poling (KDP), Two-stroke Glide (G2) and Five-stroke Glide (G5)) reached 99.5%, and the accuracy was significantly improved compared with similar recognition systems. Therefore, the equipment is expected to be a valuable training tool for coaches and athletes, helping them to better understand and improve their ski maneuver technique.

Human Trafficking Awareness Training and Response Procedure Template for Transit Agencies

As the frontline eyes on communities, transit operators are uniquely positioned to recognize suspicious activities and report them to authorities. Training is necessary to inform and empower transit agencies in the fight against human trafficking. Human trafficking is modern day slavery, which the U.S. Department of Justice defines as a crime that involves the exploitation of a person for labor, services, or commercial sex. Human trafficking is a worldwide criminal industry that the denies freedom of 25 million people around the world. This heinous crime often targets the most vulnerable populations and preys on their needs, weaknesses, and insecurities as a strategy to control their victims through force, fraud, and coercion. This paper reviews the resources available for transit agencies to train their employees on the signs of human trafficking that they should be on the lookout for. Truckers Against Trafficking’s Busing on the Lookout program has shared success stories related to transit employees recognizing situations and saving human trafficking victims. However, the existence of transit agency policies and associated procedures is a recognized resource gap in the transit industry. The available training materials direct trainees to refer to their agency policy on how to report suspicious activity. However, as noted, many transit agencies do not have a formalized policy or procedures in place to dictate how to report suspected human trafficking activity. This paper provides agencies with valuable human trafficking awareness training resources and a response procedure template as tools to improve the awareness and reporting procedures in their agency.

Validation of A 3D-printed simulator for training in endoscopic injection of bulking agent for vesicoureteral reflux: a pilot study

BackgroundSimulation-based training plays a significant role in surgical education, especially in minimally invasive pediatric surgery and urology. This study aimed to evaluate a novel 3D-printed model as training tool for endoscopic injection of bulking agent.MethodsForty-three attendees and ten teaching faculty members were invited to complete a post hoc questionnaire after completing training sessions using the Fish Tank Simulation Model (FTSM). The survey consisted of a 7-question 5-point Likert scale to assess the model’s realism (face validity) and its effectiveness as training tool (content validity).ResultsRegarding the training status, 20/53 (37.7%) participants were fellow and/or specialist in pediatric surgery and 33/53 (62.3%) were surgeons in training. Their level of confidence in endoscopic injection procedure was defined as novice (< 10 procedures per year) in 33/53 (62.3%), intermediate (10–20 procedures per year) in 10/53 (18.9%), and expert (> 20 procedures per year) in 10/53 (18.9%). Regarding both face validity and content validity assessments, no statistically significant differences were found between scores given by novice vs intermediate/expert groups. Similarly, no statistically significant differences emerged between scores given by participant vs faculty groups assessing the content validity of the FTSM. The FTSM was considered a good teaching tool for beginners by 44/53 (83%) and for pediatric surgeons/urologists by 38/53 (71.7%).ConclusionsThe 3D-printed Fish Tank Simulation Model proved to be a valuable, high-fidelity, easily accessible, cost-effective, hygienic, and domestic-use training tool for pediatric surgeons/urologists conducting the procedure. The model’s user-friendly design and realistic environment enhanced learning opportunities for trainees, regardless of their experience level or training status. Nevertheless, further development is necessary, particularly in enhancing the realism of the ureteral hiatus and reproducing more complex anatomy, to make it beneficial for the training of advanced surgeons.

Examining Professional Development for Special Education Teachers, General Education Teachers, and Paraeducators Who Support Autistic Elementary Students

ABSTRACT Background Autistic youth primarily receive services in public schools from educators who persistently report variable training and use of evidenced-based practices (EBPs). Educational achievement of individualized goals may be hindered by inconsistent implementation of EBPs. Objective To better understand EBP training practices, this study employs both quantitative and qualitative methods to examine EBP training experiences accessed by educators of different roles. Method Eighty-one participants (special education teachers, general education teachers, and paraeducators) serving autistic children in public schools completed an exhaustive survey on professional development, as well as a complementary semi-structured interview. Results Special education and general education teachers had similar patterns in terms of the types of training and providers they accessed. Compared to paraeducators, general and special educators were more likely to report accessing Teacher Preparation or Educator Certification Programs and School or District In-Service. The type of provider conducting the training also differed by role, with general educators being more likely to receive training from Higher Education Faculty, District Coaches, and Supervisors, and paraeducators more likely to receive training from Teachers. Special education teachers, general education teachers, and paraeducators also reported qualitative differences in their training experiences and needs. Conclusions This study provides valuable EBP training suggestions synthesized from the perspectives of educators working together to support autistic students. One recommendation is the facilitation of opportunities for educators to train and work together across roles, as well as the personalization of these trainings for educators of different roles, fostering a more comprehensive understanding of the needs of autistic students and promoting consistent EBP utilization.

A cross-sectional investigation of a mobile health clinic run by undergraduate pharmacy students providing services to underserved communities.

Mobile health clinics have been used to provide healthcare to underserved communities, especially during the COVID-19 pandemic. Student-led clinics, operated by undergraduate health students, offer valuable training while serving these populations. This cross-sectional study investigated the demographic characteristics and services provided by a mobile health clinic run by undergraduate pharmacy students, assessing its potential to reach underserved communities. The mobile health clinic operated from October 2023 to April 2024, staffed by 36 fourth-year pharmacy students. Services included cardiovascular disease risk screening and lifestyle advice. Demographic and service data were collected using an electronic primary care system and analysed with descriptive statistics. The clinic served 716 users, with a demographic breakdown of 53.2% female and 46.8% male, predominantly aged 31-60years. Users were ethnically diverse. Services provided included blood pressure (91.3%), BMI (91.3%), and diabetes risk assessments (54.9%). Many users reported low risk for smoking and alcohol consumption, but varied levels of physical activity. Referrals were made for cardiovascular disease risk and lifestyle support. The mobile health clinic effectively reached a diverse, underserved population, providing essential health services and facilitating student training. Further research is needed to evaluate the long-term impact and cost-effectiveness of such clinics, and the follow-up care for referred patients.

Unsupervised Domain Adaptation for EM Image Denoising with Invertible Networks.

Electron microscopy (EM) image denoising is critical for visualization and subsequent analysis. Despite the remarkable achievements of deep learning-based non-blind denoising methods, their performance drops significantly when domain shifts exist between the training and testing data. To address this issue, unpaired blind denoising methods have been proposed. However, these methods heavily rely on image-to-image translation and neglect the inherent characteristics of EM images, limiting their overall denoising performance. In this paper, we propose the first unsupervised domain adaptive EM image denoising method, which is grounded in the observation that EM images from similar samples share common content characteristics. Specifically, we first disentangle the content representations and the noise components from noisy images and establish a shared domain-agnostic content space via domain alignment to bridge the synthetic images (source domain) and the real images (target domain). To ensure precise domain alignment, we further incorporate domain regularization by enforcing that: the pseudo-noisy images, reconstructed using both content representations and noise components, accurately capture the characteristics of the noisy images from which the noise components originate, all while maintaining semantic consistency with the noisy images from which the content representations originate. To guarantee lossless representation decomposition and image reconstruction, we introduce disentanglement-reconstruction invertible networks. Finally, the reconstructed pseudo-noisy images, paired with their corresponding clean counterparts, serve as valuable training data for the denoising network. Extensive experiments on synthetic and real EM datasets demonstrate the superiority of our method in terms of image restoration quality and downstream neuron segmentation accuracy. Our code is publicly available at https://github.com/sydeng99/DADn.

Assessing the surgical competency of novice surgeons by using a three-dimensional heads-up display microscope.

To evaluate the surgical performance of novice surgeons operating on a three-dimensional (3D) heads-up display system compared to those using a traditional microscope (TM). Prospective study design in a private practice setting. Twenty novice surgeons with similar experiences in cataract surgery were selected. Each surgeon performed 20 phacoemulsification cataract surgeries: 10 surgeries on the 3D heads-up display microscope, and 10 surgeries using a TM system. Data were collected from a total of 400 patients operated on by 20 surgeons. Outcome measures were recorded and graded according to the International Council of Ophthalmology's Ophthalmology Surgical Competency Assessment Rubric-Phacoemulsification system. The main outcome measure was mean surgical competency scores. Overall, 400 data points were equally distributed between TM (200) and 3D (200) surgeries. The mean surgical competency scores were 60.19 (11.41) for TM surgeries and 62.99 (11.11) for 3D surgeries. 3D surgeries had significantly higher surgical competency scores than TM surgeries ( P = 0.013). The mean surgical duration for TM and 3D surgeries was 35.98 (6.02) min and 34.31 (7.12) min, respectively. 3D surgeries took significantly less time than TM surgeries ( P = 0.012). The overall mean best corrected visual acuity in the logarithm of the minimum angle of resolution units was 0.27 (0.42); in TM and 3D surgeries, it was 0.28 (0.43) units and 0.26 (0.41) units, respectively. The 3D heads-up display system enhances stereopsis in cataract surgery, making it a valuable training tool for novice surgeons in phacoemulsification procedures.

Dynamic Real-Time Production Forecasting Model for Complex Subsurface Flow Systems with Variable Length Input Sequences

Summary Production time-series forecasting for newly drilled wells or those with limited flow and pressure historical data poses a significant challenge, and this problem is exacerbated by the complexities and uncertainties encountered in fractured subsurface systems. While many existing models rely on static features for prediction, the production data progressively offer more informative insights as production unfolds. Leveraging ongoing production data can enhance forecasting accuracy over time. However, effectively integrating the production stream data presents significant model training and updating complexities. We propose two innovative methods to address this challenge: masked recurrent alignment (MRA) and masked encoding decoding (MED). These methods enable the model to continually update its predictions based on historical data. In addition, by incorporating sequence padding and masking, our model can handle inputs of varying lengths without trimming, thereby avoiding the potential loss of valuable training samples. We implement these models with gated recurrent unit (GRU) and evaluate their performance in a case study involving 6,154 shale gas wells in the Central Montney Region. The data set encompasses 39 production-related features, including reservoir properties, completion, and wellhead information. Performance evaluation is based on root mean square error (RMSE) to predict 36-month production from 200 wells during testing. Empirical findings highlight the efficacy of the proposed models in handling challenges associated with variable-length input sequences, showcasing their superior performance. Our research emphasizes the value of including shorter time-series segments, often overlooked, to improve predictive accuracy, especially in scenarios with limited training samples.