Multiple Experts Research Articles

Land Cover Mapping From Multiple Complementary Experts Under Heavy Class Imbalance

Land Cover Mapping From Multiple Complementary Experts Under Heavy Class Imbalance

Role of Cytology for Diagnosis of Appendageal Tumours and Comparison with Histopathology: A Cross-sectional Study

Introduction: Skin adnexal neoplasms are quite rare in routine pathology practice, yet the domain of Fine Needle Aspiration Cytology (FNAC) remains relatively unexplored. This current study explored different Adnexal Tumours (ATs) through FNAC, emphasising the efficacy of cytological diagnosis as a safe, simple, quick, and cost-effective tool with high sensitivity and specificity for diagnosing skin ATs. Aim: To diagnose various appendageal tumours through cytological examination, confirm and categorise them (benign/ malignant) via histopathological examination, and identify any cytohistological discrepancies. Materials and Methods: This retrospective cross-sectional study was conducted over a five-year period (January 2016- December 2021) at the Department of Pathology, Nil Ratan Sircar Medical College and Hospital (NRSMCH) in Kolkata, West Bengal, India, following ethical approval. Patients underwent history and physical examination followed by FNAC. ATs were diagnosed cytologically in 78 out of 23,852 FNAC cases. The results were compared with histopathological diagnosis to assess concordance or discordance based on predefined criteria. Leishman-Giemsa (L&G) stain was used for cytological examination, and Haematoxylin and Eosin (H&E) stain for histopathology. Results: Cytologically, ATs were diagnosed in 78 cases (0.33%), with the majority being benign 69 (88.46%). There was a female preponderance (49, 62.82%) with a male-to-female ratio of 1:1.7. Nodular Hidradenoma (NH) was the most common diagnosis (16/48=33.33%). Out of 59 biopsy-confirmed lesions, 55 showed cytohistological concordance (93.22%), with only four discordant cases (6.78%). The sensitivity, specificity, positive predictive value, and negative predictive value for detecting malignancy in this series were 70%, 97.90%, 87.50%, and 94.10%, respectively. Conclusion: FNAC should be the primary choice for detecting skin ATs due to its safety, reliability, high predictive value, and low chance of discordance. Although occasional cytohistological discrepancies may occur, proper clinical correlation, aspiration from multiple sites, expert cytopathologists, and high-quality staining are essential to avoid misdiagnosis.

Studying usability of public health surveillance maps through framework based heuristic evaluation

Public health surveillance systems play a crucial role in detecting and responding to disease outbreaks. Visualizations of surveillance data are important for decision-making, but little attention has been paid to the usability and interaction of such systems. In this paper, we developed a set of 10 heuristics to assess the visualization and usability of public health surveillance systems. The heuristics cover aspects of perception, cognition, and interaction. The perception deals with how the system looks in the first glance and whether it has pleasant effect on the user or otherwise. Cognition deals with the question of whether enough information is provided to use the system, while usability and interaction deal with whether the system is user-friendly in terms of the tools provided for interaction and use. We recruited a panel of experts to evaluate a set of systems using our heuristics. Results showed that there was variation in the scores of the experts' assessments, indicating the importance of multiple expert evaluations. Our heuristics provide a practical and comprehensive tool for assessing the visualization and usability of public health surveillance systems, which can lead to improved decision-making and ultimately better public health outcomes. The results suggest that the heuristic based evaluation through a panel of experts can provide meaningful results and insights into the usability aspects of public health systems. The results suggest that for some systems there can be agreement in terms of evaluation while for some other systems the experts’ opinions can vary based on the weightage and importance each expert gives to a particular aspect.

Anticoagulation Stewardship to Bridge the Implementation Gap in Perioperative Anticoagulation Management.

Lack of alignment of care protocols among providers in health care is a driver of increased costs and suboptimal patient outcomes. Perioperative anticoagulation management is a good example of a complex area where protocol creation is a clinical challenge that demands input from multiple experts. Questions regarding the need for anticoagulation interruptions are frequent. Yet, due to layers of complexity involving analysis of anticoagulation indication, surgical risk, and anesthesia-associated bleeding risk as well as institutional practices, there is heterogeneity in how these interruptions are approached. The recent perioperative anticoagulation guidelines from the American College of Chest Physicians summarize extensive evidence for the management of anticoagulant and antiplatelet medications in patients who undergo elective interventions. However, implementation of these guidelines by individual clinicians is highly varied and often does not follow the best available clinical evidence. Against this background, anticoagulation stewardship units, which exist to improve safety and quality monitoring for the anticoagulated patient, are of growing interest. These units provide a bridge for the implementation of value-based, high-quality guidelines for patients who need perioperative anticoagulation interruption. We use a case to pragmatically illustrate the problem and tactics for change management and implementation science that may facilitate the adoption of perioperative anticoagulation guidelines.

Consensus and Risk Aversion Learning in Ensemble of Multiple Experts for Long-Tailed Classification

Consensus and Risk Aversion Learning in Ensemble of Multiple Experts for Long-Tailed Classification

Multimodal Gated Mixture of Experts Using Whole Slide Image and Flow Cytometry for Multiple Instance Learning Classification of Lymphoma

In this study, we present a deep-learning-based multimodal classification method for lymphoma diagnosis in digital pathology, which utilizes a whole slide image (WSI) as the primary image data and flow cytometry (FCM) data as auxiliary information. In pathological diagnosis of malignant lymphoma, FCM serves as valuable auxiliary information during the diagnosis process, offering useful insights into predicting the major class (superclass) of subtypes. By incorporating both images and FCM data into the classification process, we can develop a method that mimics the diagnostic process of pathologists, enhancing the explainability. In order to incorporate the hierarchical structure between superclasses and their subclasses, the proposed method utilizes a network structure that effectively combines the mixture of experts (MoE) and multiple instance learning (MIL) techniques, where MIL is widely recognized for its effectiveness in handling WSIs in digital pathology. The MoE network in the proposed method consists of a gating network for superclass classification and multiple expert networks for (sub)class classification, specialized for each superclass. To evaluate the effectiveness of our method, we conducted experiments involving a six-class classification task using 600 lymphoma cases. The proposed method achieved a classification accuracy of 72.3%, surpassing the 69.5% obtained through the straightforward combination of FCM and images, as well as the 70.2% achieved by the method using only images. Moreover, the combination of multiple weights in the MoE and MIL allows for the visualization of specific cellular and tumor regions, resulting in a highly explanatory model that cannot be attained with conventional methods. It is anticipated that by targeting a larger number of classes and increasing the number of expert networks, the proposed method could be effectively applied to the real problem of lymphoma diagnosis.

Impact of low pharyngeal/esophageal pressure associated with sarcopenia on postendoscopic submucosal dissection pneumonia in patients with superficial esophageal cancer.

The aging population, including patients with superficial esophageal cancer, encounters critical dysphagia- and postoperative pneumonia-related issues. Although endoscopic submucosal dissection (ESD) provides advantages over other modalities, older patients are at higher risk of postoperative pneumonia. Furthermore, the etiologies of pneumonia are complex and include patient- (such as sarcopenia) and treatment- (including ESD) related factors. Therefore, this study evaluated swallowing function in patients with superficial esophageal cancer and identified post-ESD pneumonia-associated factors. Comprehensive swallowing function and sarcopenia were evaluated in patients pre-ESD and 2 months post-ESD using high-resolution manometry and several swallowing studies by multiple experts. The effects of mucosal resection and sarcopenia on swallowing function changes post-ESD, the relationship between preoperative swallowing function and sarcopenia, and the factors influencing postoperative pneumonia were investigated. Twenty patients were included in the study. Patients with preoperative sarcopenia had significantly lower pharyngeal/upper esophageal sphincter and tongue pressures than those without sarcopenia. However, ESD did not worsen pharyngeal or upper esophageal pressure. Post-ESD pneumonia incidence tended to be higher in patients with sarcopenia than in those without sarcopenia. The lower upper esophageal sphincter-integrated relaxation pressure (UES-IRP) was a significant factor in pneumonia development. Furthermore, the receiver operating characteristic curve for UES-IRP in pneumonia yielded an area under the curve of 0.82. Sarcopenia is associated with preoperative dysphagia, which increases post-ESD pneumonia risk. Therefore, postoperative pneumonia incidence is expected to increase with an aging population, making preoperative sarcopenia and swallowing function evaluation crucial.

The Belgian Integrated DEpression Care (IDECA) project: implementation and scientific evaluation protocol

Background: Depression creates a substantial personal burden for affected individuals and their families, and produces significant economic and social hardships that affect society as a whole. In Belgium, depression is a leading cause of disability, measured as disability-adjusted life years. The COVID-19 pandemic has greatly exacerbated the already substantial health and socioeconomic consequences of depression and has led to major changes in the need for and delivery of mental health services. Integrating mental health services into primary care is the most viable way to closing existing treatment gaps. The majority of people with mental disorders treated in primary care have good outcomes, particularly when linked to a network of services at secondary level and in the community. A balanced model of depression care calls for a system with community-based mental health care, delivered by primary care physicians and supported by other professionals in an eco-system of care.
 Aim: Implementation of an integrated depression care (IDECA) intervention in two regions in Flanders, Belgium with evaluation of the intervention’s working components, its acceptance by patients and care providers, and the contextual conditions which needs to be fulfilled to safeguard success of the model of care.
 Methods: A scientific steering committee including multiple academic expertise areas and patient representatives guided the development of the intervention and its scientific evaluation. Moreover, feedback was obtained from the Flemish Institute for Primary Care (VIVEL), the Family Doctors Medical Association (Domus Medica), and a mental health interest group(Psyche VZW). This resulted in an implementation design without control group and a targeted heterogeneous patient population with depression to obtain real-world evidence. IDECA consists of three intervention pillars: (i) the integration of two reference persons mental health (case management function) into general medical practice and who provide diagnostic support, coordination, follow up care, and patient education; (ii) the implementation of a validated shared care treatment protocol; and (iii) professional training on depression screening and management. The three pillars will be further operationalized in co-creation with the selected regions. The intervention will run for 15 months.
 Data will be collected from both participating care providers and patients. Evaluation will take place using a Realist Evaluation approach in combination with the NoMad questionnaire for care providers which measures the four implementation metrics of the Normalization Process Theory: i. coherence, ii. cognitive participation, iii. collective action, and iv. reflexive monitoring. NoMad process data will serve as a basis for semi-structured qualitative research on Context-Mechanism-Outcome constellations. Questionnaires will be used to measure depression and anxiety levels, health literacy, quality of life, patient satisfaction, therapy adherence and self-efficacy, and healthcare resource utilization in patients.
 Results: Patient enrollment will start at the beginning of 2023. Preliminary data on patient characteristics and first data on care providers’ acceptance and engagement with the IDECA project will be presented, next to the co-creative development process of the intervention.
 Conclusion: The IDECA-study aims at the implementation of integrated depression care and evaluation of its feasibility in two different regions in Belgium.

Maximizing Multiple Intelligences for Unprecedented Academic Achievement

This study explores the integration of multiple intelligences and enhancing student motivation in Indonesian classrooms. Individual diversity is respected, inclusive education is promoted, student engagement is maximized, and academic motivation is increased. Educators may develop engaging learning environments that respond to students' talents and interests by identifying and integrating multiple intelligences. To collect perspectives from educators, students, and stakeholders, the study used qualitative approaches such as interviews and observations. According to the findings, using multiple intelligences improves student engagement and academic accomplishment. However, issues such as teacher training and assessment systems must be addressed. Collaboration among educators, policymakers, and stakeholders is critical for implementation success. Indonesia can develop a dynamic and empowering educational system that releases students' full potential by integrating multiple intelligences and local expertise.

Group decision making in the analytic hierarchy process by hesitant fuzzy numbers

Due to the increasing complexity of decision problems, many managers employ multiple experts to reach a good decision in a group decision making. Now, if there is ambiguity in the evaluation of experts, the use of fuzzy numbers can be useful for each expert. In these situations, the use of hesitant fuzzy numbers (HFNs) which consists of several fuzzy numbers with special conditions can be suggested. HFNs are as an extension of the fuzzy numbers to take a better determining the membership functions of the parameters by several experts. Because of simple and fast calculations, in this paper, we use triangular HFNs in the pairwise comparison matrix of analytic hierarchy process by opinions of a group of decision makers in a hesitant fuzzy environment. We define consistency of the hesitant fuzzy pairwise comparison matrix and use the arithmetic operations on the HFNs and a new method of comparing HFNs to get the hesitant fuzzy performance score. By using score function to hesitant fuzzy score we can get a final score for alternatives. Finally, a practical example is provided to show the the effectiveness of this study. The obtained results from this paper show that new method can get a better answer by keeping the experts’ opinions in the process of solving the problem.

OP141 Expert Knowledge Elicitation in Health Technology Assessment: Our Experience Using the Sheffield Elicitation Framework

IntroductionExpert judgement has an important role in health technology assessment (HTA), including as a source of evidence to inform economic modeling when published data are lacking. Quantitative information may be elicited from experts to inform model inputs and associated uncertainty using one of many expert elicitation methodologies. Here, the feasibility and potential benefits of one expert elicitation method, the Sheffield Elicitation Framework (SHELF), to the HTA process is examined.MethodsThe SHELF method seeks to express the knowledge of multiple experts in the form of a subjective probability distribution. Eliciting a subjective probability distribution allows the uncertainty of experts to be included in probabilistic sensitivity analysis, which is becoming an increasingly prominent feature of HTAs. The individual knowledge of participating experts is combined through behavioral aggregation, where experts participate in a discussion before being asked to provide judgments from the perspective of a rational impartial observer. The whole process is led by a facilitator who ensures all participants contribute and confirm that the final distribution is a product of consensus, not compromise.ResultsWe recently conducted two SHELF elicitations as part of an ongoing project aiming to streamline the assessment of positron emission tomography (PET) in Australia. These elicitations provided insight into the usefulness of SHELF within the HTA setting. Given the constraints imposed by the COVID-19 pandemic, the elicitation sessions were conducted online rather than in the ideal face-to-face manner. In collaboration with one of the developers, we successfully adapted the method by making use of video conferencing technology to provide an online environment that mimicked the face-to-face setup as much as possible.ConclusionsSHELF provides a rigorous and scientific method by which to elicit the knowledge of multiple experts in the form of a probability distribution. However, the method is resource intensive and may be best reserved for when data on key drivers are lacking.

Unveiling decomposition dynamics: leveraging 3D models for advanced forensic analysis.

Forensic taphonomy, the study of post-mortem processes, is pivotal in modern forensic science. This short communication illuminates limitations in traditional 2D imaging, specifically digital photographs, within forensic taphonomy, and highlights the vast potential of 3D modeling techniques. Drawing from a recent study in Hawaii's tropical savanna, we unveil disparities between real-time observations and 2D photographs when assessing decomposition, emphasizing the importance of scoring method selection and the need to scrutinize 2D imaging's accuracy in forensic taphonomy. Conversely, 3D modeling techniques, an emerging powerhouse in forensic science, offer multidimensional data, including volume, surface area, and spatial relationships, allowing for comprehensive and precise representation of decomposition dynamics. Despite concerns about texture quality, 3D models yield objective data amenable to analysis by multiple experts, thus minimizing subjectivity and augmenting the reliability of forensic assessments. The potential for 3D modeling to bridge the gap between 2D imaging and real-time decomposition requires tailored methodologies. Future research should focus on standardizing protocols and fostering collaboration among forensic experts, technologists, and researchers to unleash 3D technology's full potential in advancing forensic taphonomy.

3D vs. 2D simulated fetoscopy for spina bifida repair: a quantitative motion analysis

3D imaging technology is becoming more prominent every day. However, more validation is needed to understand the actual benefit of 3D versus conventional 2D vision. This work quantitatively investigates whether experts benefit from 3D vision during minimally invasive fetoscopic spina bifida (fSB) repair. A superiority study was designed involving one expert team (>18 procedures prior) who performed six 2D and six 3D fSB repair simulations in a high-fidelity animal training model, using 3-port access. The 6D motion of the instruments was recorded. Among the motion metrics are total path length, smoothness, maximum speed, the modified Spectral Arc Length (SPARC), and Log Dimensionless Jerk (LDLJ). The primary clinical outcome is operation time (power 90%, 5% significance) using Sealed Envelope Ltd. 2012. Secondary clinical outcomes are water tightness of the repair, CO_2 insufflation volume, and OSATS score. Findings show that total path length and LDLJ are considerably different. Operation time during 3D vision was found to be significantly shorter compared to 2D vision (113pm 9 vs. 149pm 19 min; p = 0.026). These results suggest enhanced performance with 3D vision during interrupted suturing in fetoscopic SBA repair. To confirm these results, a larger-scale follow-up study involving multiple experts and novice surgeons is recommended.

A q-rung orthopair fuzzy multi-attribute group decision making model based on attribute reduction and evidential reasoning methodology

Multi-attribute group decision-making (MAGDM) is an analytical decision-making framework which gathers the views of multiple experts, and they often assess alternatives by degree adverbs in realistic scenarios. In order to treat expert opinions well, q-rung orthopair fuzzy sets (q-ROF) emerged as a data disposal tool, however, how to aggregate the information rationally remains a challenge. In this paper, q-ROF evidential reasoning methodology (q-ROF ERM) is proposed to guarantee the functionality in the case of conflicting, extreme information evaluations from experts and attributes. Second, the q-ROF attribute reduction (AR) framework is invented for attribute importance measurement. The flexibility and elegance of the model are boosted by the variation of the λ parameter and its lead to the elimination of redundant attributes. In addition to this, the q-ROF output three-way decision (TWD) framework is created to give the classification and ranking of alternatives. The relative output function derived from fuzzy distance measure and the adjustable utility pursuit coefficient τ afford both objectivity and subjectivity to the program selection, allowing the evaluation outcomes to be relevant to human cognition. Finally, the comparative experiments and sensitivity analyses of the two cases illustrate the scientificity and reliability of the presented framework in the MAGDM scenario.

Opportunistic identification of atrial fibrillation during unscheduled contact with emergency ambulance services in England: a qualitative exploration

Abstract Background Atrial fibrillation (AF) is an important risk factor for stroke which commonly affects people attended by Emergency Medical Services (EMS). There may be an opportunity for. EMS clinicians to find new incidental cases of AF suitable for oral anticoagulation (OAC) amongst the population who do not require conveyance to hospital following urgent ambulance review. Purpose This study explored whether it was acceptable to propose a formal pathway for identification and subsequent referral of non-conveyed EMS patients with incidental AF to their general practitioner (GP) for consideration of OAC prescription. The perspectives of patients, paramedics, and field experts were included. Methods This qualitative study included four online focus groups involving 18 members of the public and one-to-one online interviews in England with 11 health care and service providers (6 paramedics, 5 field experts providing multiple expertise in cardiology, general practice, public health, academic research, healthcare policy, and primary care commissioning) between June and September 2022. A semi-structured topic guide was used, with purposive sampling to maximise the diversity of perspective and experience in all participant groups. Data were audio-recorded, transcribed anonymously and analysed thematically. Results All participant groups supported formalising the opportunistic detection of AF during EMS review, but each had concerns and caveats. Of paramount concern to all groups was that any change in practice must not create delays in the emergency response system since incidental AF is non-urgent and other community pathways are in place for detection and management. Public participants were also concerned about how the diagnosis would be communicated to the patient, and opinion was split between whether this should be ‘on the scene’ or in a subsequent GP appointment. Paramedics reported that they frequently identify incidental AF, but that there is variation in practice regarding whether (and how) this is communicated to the GP. There was a concern that the finding of AF was not easily visible within the existing EMS report and may be missed. Paramedics also focussed on safety-netting of patients not being conveyed to hospital and of a need for an ‘active’ reporting process to ensure that the finding of AF was acted upon. In general, field experts felt that a formal pathway would be useful but relatively low impact, and favoured a simple intervention that did not overly complicate existing AF pathways. Conclusions There is support amongst the public and professionals for the development of a formal pathway to ensure ongoing follow-up and care for people with AF incidentally detected in the community by EMS. This novel case finding opportunity has the potential to improve AF detection, oral anticoagulation rates, and reduce the risk of stroke.

Framework for Radiation Oncology Department-wide Evaluation and Implementation of Commercial Artificial Intelligence Autocontouring

IntroductionArtificial intelligence (AI) based auto-contouring in radiation oncology has potential benefits such as standardization and time savings. However, commercial AI solutions require careful evaluation prior to clinical integration. We developed a multidimensional evaluation method to test pre-trained AI-automated contouring solutions across a network of clinics. MethodsCurated data included 121 patient planning CT (computed tomography) scans with a total of 859 clinically approved contours used for treatment from four clinics. Regions of interest (ROIs) were generated with three commercial AI-based automated contouring software solutions (AI1, AI2, AI3) spanning the following disease sites: brain, head-and-neck, thorax, abdomen, and pelvis. Quantitative agreement between AI-generated and clinical contours was measured by Dice similarity coefficient (DSC) and Hausdorff distance (HD). Qualitative assessment was performed by multiple experts scoring blinded AI-contours using a Likert scale. Workflow and usability surveying was also conducted. ResultsAI1/AI2/AI3 contours had high quantitative agreement in 27.8/32.8/34.1% of cases (DSC>0.9), performing well in pelvis (median DSC = 0.86/0.88/0.91) and thorax (median DSC = 0.91/0.89/0.91). All three solutions had low quantitative agreement in 7.4/8.8/6.1% of cases (DSC<0.5), performing worse in brain (median DSC=0.65/0.78/0.75) and H&N (median DSC=0.76/0.80/0.81). Qualitatively, AI1/AI2 contours were acceptable (rated 1-2) with at most minor edits in 70.7/74.6% of ROIs (2,906 ratings), higher for abdomen (AI1: 79.2%) and thorax (AI2: 90.2%), and lower for H&N (29.0/35.6%). An end-user survey showed strong user preference for full automation and mixed preferences for accuracy versus total number of structures generated. ConclusionOur evaluation method provided a comprehensive analysis of both quantitative and qualitative measures of commercially available pre-trained AI auto-contouring algorithms. The evaluation framework served as a roadmap for clinical integration that aligned with user workflow preference.

Implementation of a national cancer subspecialist virtual consult pilot in an integrated health system.

143 Background: Over 65,000 new cancer patients are diagnosed nationally each year at Kaiser Permanente (KP), with 12.7 million members in 9 states and the District of Columbia. KP developed the Cancer Expert Review Program (the Program) to connect oncologists with cancer subspecialists across the enterprise. The Program provides consultation and a chart note viewable by the patient within 2 business days. Methods: The Program was piloted from October 2022 to June 2023 for medical oncology patients with breast, gastrointestinal (GI), genitourinary (GU), sarcoma, thoracic, hematology, central nervous system (CNS), melanoma, head &amp; neck, and gynecologic oncology (GYN) diagnoses or a need for molecular/genomic testing. Medical oncologists from Hawaii, Georgia, and Southern California submitted requests for virtual consults to designated cancer subspecialists using HealthConnect, KP’s EPIC system. Consulting physicians had direct access to the medical records for each patient, including pathology and imaging. Oncologists informed patients of the service and discussed the rendered opinion as part of shared decision-making. Implementation outcomes of feasibility and acceptability were evaluated. Providers were surveyed about perceptions of the service using an internally developed questionnaire. The time from consult request to completion was measured. Results: 108 consults (67.5%) were provided by individual subspecialists, while 22 consults (14%) were reviewed by multiple experts or a multidisciplinary team. 140 consults (87.5%) were completed within 2 business days. Of the 44 oncologists who requested a consult, 12 responded to a physician survey. 85% of physicians rated satisfaction with the service as “satisfied” or “extremely satisfied”. 75% said the advice helped to develop the treatment plan. Half of physicians reported the advice changed the original treatment or approach either “somewhat” or “to a great extent.” Half of physicians indicated that if this internal program were not available, they would have referred their patient to a non-Kaiser Permanente provider for a consultation. Conclusions: Implementation of a national cancer subspecialist virtual consult pilot within an integrated health system demonstrated the feasibility and acceptability of timely expert consults with high physician satisfaction. The Program facilitates subspecialist care in collaboration with the patient’s trusted general oncologist, keeping care close to home.Additional research is needed to understand the long-term implementation and quality outcomes of the Program.[Table: see text]

The Selection of Investment Priorities using Expanded Dual Hesitant Fuzzy Sets

The concept of expanded dual hesitant fuzzy set is a concept that can be applied in decision-making problems. In decision-making problems, expanded dual hesitant fuzzy set can be applied to represent the opinions of multiple experts or stakeholders in a more elaborate manner. It enables decision-makers to provide more detailed information about their preferences and hesitancy.

Microsurgical anastomosis of the fallopian tubes after tubal ligation: a systematic review and meta-analysis

ObjectiveBetween 20% and 30% of women who have undergone tubal ligation regret their decision. The alternative to regain fertility for these women is either in vitro fertilization or tubal re-anastomosis. This article presents a systematic review with meta-analysis to assess the current evidence on the efficacy of tubal recanalization surgery in patients who have previously undergone tubal ligation. Study designThe search was conducted in the World of Science (WOS) database, The Cochrane Library and ClinicalTrials.gov record using the keywords “tubal reversal”, “tubal reanastomosis” and “tubal anastomosis”. The review was carried out by two of the authors. Data from 22 studies were evaluated, comprising over 14,113 patients who underwent the studied surgery, following strict inclusion criteria: articles published between January 2012 and June 2022, in English and with a sample size bigger than 10 patients were included. A random-effects meta-analysis was performed. ResultsThe overall pregnancy rate after anastomosis was found to be 65.3 % (95 % CI: 61.0–69.6). The percentage of women who had at least one live birth, known as the birth rate, was 42.6 % (95 % CI: 34.9–51.4). Adverse outcomes after surgery were also examined: the observed abortion rate among women who underwent surgery was 9.4 % (95 % CI: 7.0–11.7), and the overall ectopic pregnancy rate was 6.8 % (95 % CI: 4.6–9.0). No differences were found between the outcomes when differentiating surgical approaches: laparotomy, laparoscopy, or robotic-assisted surgery. The patient's age was identified as the most significant determining factor for fertility restoration. Finally, when comparing the results of tubal reversal with in vitro fertilization, reversal procedures appear more favorable for patients over 35 years old, while the results are similar for patients under 35 years old, but more data is needed to evaluate this finding. ConclusionTherefore, the available literature review demonstrates that surgical anastomosis following tubal ligation is a reproducible technique with relevant success rates, performed by multiple expert groups worldwide.

NCL++: Nested Collaborative Learning for long-tailed visual recognition

Long-tailed visual recognition has received increasing attention in recent years. Due to the extremely imbalanced data distribution in long-tailed learning, the learning process shows great uncertainties. For example, the predictions of different experts on the same image vary remarkably despite the same training settings. To alleviate the uncertainty, we propose a Nested Collaborative Learning (NCL++) which tackles the long-tailed learning problem by a collaborative learning. To be specific, the collaborative learning consists of two folds, namely inter-expert collaborative learning (InterCL) and intra-expert collaborative learning (IntraCL). InterCL learns multiple experts collaboratively and concurrently, aiming to transfer the knowledge among different experts. IntraCL is similar to InterCL, but it aims to conduct the collaborative learning on multiple augmented copies of the same image within the single expert. To achieve the collaborative learning in long-tailed learning, the balanced online distillation is proposed to force the consistent predictions among different experts and augmented copies, which reduces the learning uncertainties. Moreover, in order to improve the meticulous distinguishing ability on the confusing categories, we further propose a Hard Category Mining (HCM), which selects the negative categories with high predicted scores as the hard categories. Then, the collaborative learning is formulated in a nested way, in which the learning is conducted on not just all categories from a full perspective but some hard categories from a partial perspective. Extensive experiments manifest the superiority of our method with outperforming the state-of-the-art whether with using a single model or an ensemble. The code will be publicly released.