Valid Comparisons Research Articles

Benchmarking in inflammatory bowel disease: lessons from Australia and New Zealand

Abstract Inflammatory bowel disease (IBD) is a complex chronic disease estimated to affect almost 1% of the Western population, where proactive care has largely replaced a reactive approach. Whilst there have also been great strides in new care models and therapies, care quality, and outcomes are poorly known, leaving us designing and delivering services without an effective feedback loop. Quality of care, guidelines, consensus statements, and standards will each be discussed, as a solid understanding of these concepts is essential. Assessing quality involves measuring the 3 core domains of structure, process, and outcomes. Benchmarking is an extension of quality assessment, which is recurrent or continuous and enables transparent, valid comparisons of quality between care sites. Assessment of QUALITY and BENCHMARKING are resource intensive unless processes are automated and use routinely collected data generated in the moments of care. Digital tools now provide realistic opportunities to make this affordable and sustainable, though concerns around privacy and security hamper solutions, whilst siloed data within institutions restrict comparative use. One novel solution to enable IBD benchmarking has been developed in Australia and New Zealand (ANZ). Crohn’s Colitis Care (CCCare) is a cloud-based, IBD-specific EMR, now being used in routine care documentation at 22 sites across ANZ. De-identified data feed from CCCare’s clinical module into a Clinical Quality Registry, in which audit, research, and benchmarking can be performed. The pathway from conception to implementation will be presented. Lessons learnt from the ANZ benchmarking journey will be discussed and suggestions for an effective global approach to IBD data will be proposed.

Social Disengagement in Post-Pandemic China: A Translation, Validation and Cross-Cultural Comparison of the Pandemic Disengagement Syndrome Scale.

In contrast to abundant research on the various acute mental effects of COVID-19, the long-term influences of the pandemic are still underexplored in China owing to the paucity of assessment tools. The Pandemic Disengagement Syndrome Scale (PDSS) assesses people's social disengagement as a lasting psychological consequence in Western countries during the post-COVID-19 pandemic era. However, its generalizability across cultures is untested. The present studies aimed to validate Chinese PDSS and compare disengagement syndrome levels among China, the United States and Italy. In Study 1, a Chinese version of the PDSS was developed, psychometric properties including factor structure, internal consistency, measurement invariance across gender and country, discriminant validity, and test-retest reliability were tested. Study 2 examined demographic differences in the pandemic disengagement syndrome in China and compared Chinese PDSS scores and those in the United States and Italy (Ns = 415US, 455Italy, 826China). The findings indicated that disengagement syndrome may exist among Chinese people even substantially after the acute phases of the pandemic. Meanwhile, the Chinese PDSS demonstrating acceptable psychometric features can be a valid instrument to assess the syndrome. Several possible reasons for the persistence of disengagement in China are discussed.

Ideological religious nationalism: measurement, construct validity, and cross-cultural comparisons

ABSTRACT Ideological religious nationalism (IRN) is a worldview that advocates the integration of religious beliefs with national policy and laws and the religious moralization of politics. However, the psychological mechanisms, individual differences, and socio-political consequences related to IRN are unclear. Across five studies (N Total = 1,349), we established construct validity for a novel scale assessing IRN that is adaptable across different contexts. Results showed that stronger IRN relates to distinct psychological motives, domains of religiosity, and views of one’s nation. Results also showed that stronger IRN relates to supporting policies that promote religious-national integration and support of political violence. Lastly, results show that this novel IRN scale can be used across different contexts and has incremental validity beyond similar but distinct measures of religious nationalism. Overall, results showed that IRN can successfully capture people’s integration of their religious beliefs with their views about their nation’s identity.

AMSformer: A Transformer for Grain Storage Temperature Prediction Using Adaptive Multi-Scale Feature Fusion

Grain storage temperature prediction is crucial for silo safety and can effectively prevent mold and mildew caused by increasing grain temperature and condensation due to decreasing grain temperature. However, current prediction methods lead to information redundancy when capturing temporal and spatial dependencies, which diminishes prediction accuracy. To tackle this issue, this paper introduces an adaptive multi-scale feature fusion transformer model (AMSformer). Firstly, the model utilizes the adaptive channel attention (ACA) mechanism to adjust the weights of different channels according to the input data characteristics and suppress irrelevant or redundant channels. Secondly, AMSformer employs the multi-scale attention mechanism (MSA) to more accurately capture dependencies at different time scales. Finally, the ACA and MSA layers are integrated by a hierarchical encoder (HED) to efficiently utilize adaptive multi-scale information, enhancing prediction accuracy. In this study, actual grain temperature data and six publicly available datasets are used for validation and performance comparison with nine existing models. The results demonstrate that AMSformer outperforms in 36 out of the 58 test cases, highlighting its significant advantages in prediction accuracy and efficiency.

Particle-in-cell methods in edge plasma physics: the PICLS code

Abstract Over the past decades, multiple gyrokinetic codes have shown to be able to simulate turbulence and associated transport in the core of Tokamak devices. However, their application to the edge and scrape-off layer (SOL) region presents significant challenges. To date, only few codes and models have been adapted to SOL/edge conditions. To further study the SOL region in particular, with its steep temperature and density gradients as well as large fluctuation amplitudes, the full-f particle-in-cell code PICLS has been developed. PICLS is based on a full-f gyrokinetic model with linearised field equations, considers kinetic electrons and uses logical sheath boundary conditions. In the past, PICLS was verified by applying it to a well-studied 1D parallel transport problem during an edge-localized mode in the SOL under both collisionless and collisional conditions, for which a Lenard-Bernstein collision operator was implemented. PICLS recently was extended towards three spatial dimensions to study turbulence in open-field-line regions in slab and closed-field-line toroidal geometries. In this work, we will focus on the models and methods we used for extending the code towards three spatial dimensions, including validation efforts and comparisons with other existing codes in closed-field-line geometry.

Methodological Issues in Measuring Restrictive Care Practices (Mechanical/physical restraint, Chemical restraint and Seclusion) in Adult Mental Health Inpatient Units: A Systematic Review of Recent Literature.

To identify and characterise the approaches and instruments used in recent literature to measure the prevalence of restrictive care practices in adult mental health inpatient units. Additionally, it sought to summarise the reported psychometric properties, including reliability and validity of these measures. A systematic review of recent litratures was conducted using Scopus, MEDLINE, CINAHL, PsycINFO, Web of Science and Embase databases to identify studies published from 1 January 2010 to 11 October 2023. A total of 128 studies measuring the use of restrictive care practices were included. Data on measurement methods were extracted from each study and summarised to compare how consistently these practices have been measured across studies and how authors consistently reported the reliability and validity of these measurment approaches. All findings were reported following the PRISMA 2020 checklist. There were significant variations in how the prevalence of restrictive care practices was measured, and the reliability and validity of these measurements were unclear for most studies. Only 11 studies reported inter/intra-rater reliability. Key variations were observed in data sources utilised, how and by whom the data were collected, the timing and total duration of data collection during patient admission, how and by whom data were extracted from secondary sources, measurement instruments and the reported reliability and validity of measures. Methodological inconsistencies about the measurements approaches of restricitve care practices would introduduce potential random and/or systematic biases on the reported data which may obscure the the true prevalance these practices. This hinder the abilityto acurately assess the effectiveness of reduction strategies and understand the naturally occuring practices. Establishing a standardised set of reliable measures is crucial for enabling valid comparisons for the rates of restricitve car epractice use across settings and countries, which could enhance the ongoing monitoring and reduction of these practices. The absence of standardised defintions and measurement approaches for restrictive care practices challenges the global effort to reduce their use. Without reliable and common measures, clinicians and researchers often face challenges in documening RCP incidents accurately, compromising efforts to improve care quality and support a recovery-oriented approach. Such measurment errors would mislead decission-maker which would furhter contribute to the inconsistency the the implementation of these practices. No patient or public contribution. PROSPERO: CRD:42022335167; https://www.crd.york.ac.uk/prospero/export_details_pdf.php.

The utility of Chat Generative Pre-trained Transformer as a patient resource in paediatric otolaryngology.

This study aimed to determine the feasibility of using Chat Generative Pre-trained Transformer (ChatGPT) (https://chatgpt.com) as a patient resource for paediatric otolaryngology conditions and assess the quality of responses generated by ChatGPT when compared with information available on the internet. ChatGPT responses to common paediatric otolaryngology conditions were compared with top internet pages for readability (Flesch Reading Ease score, word count), expediency (time taken to generate response), validity (comparison of recommendations to the American Academy of Otolaryngology Head and Neck Surgery guidelines) and consistency (changes in recommendations based on question variation). ChatGPT was more expeditious in generating responses with fewer words, albeit with higher reading scores. When compared with accredited guidelines, there was no difference in validity between these sources (internet sources and ChatGPT). Consistent responses were obtained with question variation. ChatGPT may be a valuable source for patients and families in providing valid information comparable to internet materials.

A Classification System of Asian Noses for Rhinoplasty.

While Asian noses are generalized as having smaller bones and softer cartilages, thicker skin and soft tissue envelop (SSTE), there exists a diversity of morphology amongst Asian noses. The oversimplification of Asian noses diminishes the validity or generalizability of rhinoplasty research findings and makes comparison of outcomes challenging. A classification system is proposed that reflects the different structures and SSTEs of the nasal subtypes, the techniques required in the management of each of the subtypes, the challenges the surgeon will face intra- and postoperatively, allows for preoperative counselling of expected outcomes, and facilitate valid and fair comparison of study outcomes by comparing like for like. The classification system stratifies noses into four subtypes based on (1) the size and strength of the nasal bone and cartilages, and (2) the thickness of the SSTE. Type I noses have thinner SSTE and stronger structure. Type II noses have thinner SSTE and weaker structure. Type III noses have thicker SSTE and stronger structure. Type IV noses have thicker SSTE and weaker structure. Interrater variability in classifying noses among rhinoplasty surgeons was found to be very high with a kappa coefficient of 0.933 (95% confidence interval: 0.852-1.014; variance 0.013).

What Is Needed to Determine and Potentially Improve the Clinical Effectiveness and Cost-Effectiveness of Robot-Assisted Reconstructive Urinary Tract and Pelvic Floor Surgery?-ICI-RS 2024.

Robot-assisted (RA) surgery has many advantages for urinary tract and pelvic floor reconstruction (UTPFR) but its application is only slowly increasing. Studies are limited in number, patients and follow-up duration. The low-volume high-complexity nature of UTPFR warrants specific research to evaluate the potential role of the RA approach. During its 2024 meeting in Bristol, the International Consultation on Incontinence-Research Society (ICI-RS) reviewed the current literature on RA UTPFR. Priority research questions and strategies were explored in a proposal. Four different topics were discussed. (1) Large multicentre prospective studies on RA UTPFR are needed to evaluate clinical effectiveness and safety. The RA approach needs to be prioritized in specific procedures and patients, to first improve patient care where gains are highest. (2) Models for cost-utility analysis for RA UTPFR are needed to (re)direct limited resources as efficiently as possible in a quickly changing technological field. (3) A modular curriculum for training UTPFR surgeons to become RA certified needs to be developed, to allow valid interpretation and comparison of outcomes. (4) Technological improvements to address tissue handling, accessibility, efficiency and sustainability are needed. Digitalization can be a gamechanger for RA UTPFR through telesurgery (training and expertize coverage), augmented reality and artificial intelligence providing intra-operative assistance, assessment and feedback. The proposal discussed the need for studies on clinical and cost effectiveness, safety, and training curriculum development for RA UTPFR. Recent technological developments hold great promise for RA UTPFR. High-priority research questions were identified.

NIMG-39. THE “RESECTABILITY INDEX” AS A NOVEL OBJECTIVE PREDICTOR OF SURVIVAL FOR PATIENTS WITH ADULT-TYPE DIFFUSE GLIOMAS

Abstract Extent of resection (EOR) is a strong prognostic factor for patients with adult-type diffuse gliomas. Whole brain tractography (WBT) allows visualization of tumors in relation to functional subcortical white matter tracts connecting eloquent cortex. We developed a preoperative metric, the resectability index (RI), to estimate tumor resectability based on pre-resection volume (PRV) and unresectable tumor volume (UTV). We explore if RI predicts progression-free survival (PFS) and overall survival (OS). We retrospectively reviewed the charts of consecutive patients with adult-type diffuse gliomas resected from 2013-2020. Only patients with pre-operative DTI MRIs were included. UTV was the total overlapping volume of tumor with functionally significant tracts (corticospinal, arcuate, optic, fornix) and deep structures (basal ganglia, thalamus, brainstem). RI was (PRV-UTV)/PRV. Seventy-nine patients were included with an average age of 56 years at time of surgery. Glioblastoma was the most common histology (68%). Tumors were located in the left hemisphere (LH) (48%), right hemisphere (RH) (39%), and bilateral hemispheres or midline (BHM) (13%). EOR for LH, RH, and BHM tumors was 0.8, 0.91, and 0.29, respectively (P = 0.1^7). RI for LH, RH, and BHM tumors was 0.86, 0.912, and 0.44, respectively (P = 0.00021). Cox regression was used to model PFS and OS based on age, preoperative KPS, WHO grade, EOR, and RI. Age, recurrent status, bilaterality, residual tumor volume (RTV) and UTV predicted PFS (CI = 1.01 – 1.06, P = 0.002). For OS, UTV and RI were more strongly predictive than RTV or EOR. In a Cox regression model consisting only of preoperative variables, RI was independently predictive of OS. RI is a novel metric with potential efficacy for preoperative survival prognostication in patients with diffuse gliomas. RI may enable neurosurgeons to provide HGG patients with more accurate information about expected benefits of surgical intervention than is currently possible. RI will also enable more valid comparisons of HGG patient cohorts treated at different institutions.

A novel maternal thyroid disease prediction using multi-scale vision transformer architecture with improved linguistic hedges neural-fuzzy classifier.

Early pregnancy thyroid function assessment in mothers is covered. The benefits of using load-specific reference ranges are well-established. We pondered whether the categorization of maternal thyroid function would change if multiple blood samples obtained early in pregnancy were used. Even though binary classification is a common goal of current disease diagnosis techniques, the data sets are small, and the outcomes are not validated. Most current approaches concentrate on model optimization, focusing less on feature engineering. The suggested method can predict increased protein binding, non-thyroid syndrome (NTIS) (simultaneous non-thyroid disease), autoimmune thyroiditis (compensated hypothyroidism), and Hashimoto's thyroiditis (primary hypothyroidism). In this paper, we develop an automatic thyroid nodule classification system using a multi-scale vision transformer and image enhancement. Graph equalization is the chosen technique for image enhancement, and in our experiments, we used neural networks with four-layer network nodes. This work presents an enhanced linguistic coverage neuro-fuzzy classifier with chosen features for thyroid disease feature selection diagnosis. The training procedure is optimized, and a multi-scale vision transformer network is employed. Each hop connection in Dense Net now has trainable weight parameters, altering the architecture. Images of thyroid nodules from 508 patients make up the data set for this article. Sets of 80% training and 20% validation and 70% training and 30% validation are created from the data. Simultaneously, we take into account how the number of training iterations, network structure, activation function of network nodes, and other factors affect the classification outcomes. According to the experimental results, the best number of training iterations is 500, the logistic function is the best activation function, and the ideal network structure is 2500-40-2-1. K-fold validation and performance comparison with previous research validate the suggested methodology's enhanced effectiveness.

Rapid Dual-Gene Detection of Burkholderia gladioli and Subspecies Cocovenenans in Fresh Noodles and Tremella Using CRISPR-Cas12a/Cas13a.

Burkholderia gladioli pv. cocovenenans is a notable foodborne pathogen that poses a significant risk to food safety. Contaminated food requires distinct classification and treatment procedures for non-pathogenic B. gladioli and its pathogenic subspecies cocovenenans. Hence, establishing a rapid and sensitive detection method to distinguish them is necessary. In this study, we aimed to establish a method combining the CRISPR-Cas12a/Cas13a dual system with recombinase aided amplification for rapid, specific and sensitive detection of non-pathogenic B. gladioli and pathogenic subspecies cocovenenans in food. First, a RAA-CRISPR-Cas12a/Cas13a method was developed and its feasibility was assessed. Next, specificity was analyzed using 23 strains of B. gladioli and 5 non-target strains. Following this, sensitivity was evaluated by preparing gradient dilutions of B. gladioli pv. cocovenenans bacterial solutions. At last, real food test samples, including fresh noodles and tremella artificially contaminated with B. gladioli pv. cocovenenans, were utilized for method validation and sensitivity comparison. The established RAA-CRISPR-Cas12a/Cas13a method exhibited high specificity and achieved 100% accuracy in detecting species B. gladioli and its subspecies cocovenenans. This rapid method could be finished within 45 min with a detection limit of 10° CFU/mL for bacterial concentration. Additionally, it achieved detection limits of 102 CFU/g for fresh noodles and 103 CFU/g for tremella. The rapid RAA-CRISPR-Cas12a/Cas13a method demonstrated high specificity and sensitivity in detecting and disgusting species B. gladioli and subspecies cocovenenans in both food test samples and post-cultivation colonies. The RAA-CRISPR-Cas12a/Cas13a method presented in this study offers a novel molecular approach for the rapid, accurate and sensitive detection of B. gladioli and its subspecies cocovenenans in foods.

Symptoms Burden as a Clinical Outcomes Assessment in HeartFailure Patients With Atrial Fibrillation.

Safe and effective pharmacologic therapy for atrial fibrillation (AF) in heart failure (HF) is an unmet need. In AF clinical trials, the standard primary endpoint of time to first symptomatic AF event (TTFSE) has several disadvantages, which could theoretically be overcome by measurement of AF-specific symptoms burden during an entire follow-up period. The authors sought to develop and validate a method of measuring symptom burden of AF in a HF population. The authors constructed a patient-reported outcome instrument (Atrial Fibrillation Symptoms Questionnaire [AFSQ]) to function in the setting of AF/HF, and used it to measure symptoms throughout long-term follow-up. The AFSQ queries the presence of 10 new or worsening symptoms, equally divided between those associated with AF or HF. In a 267 patient AF/HF trial comparing 2 AF prevention treatments, the AFSQ was linked to electrocardiography documented AF to form a 2-component clinical outcome assessment (SxBAF) that was subjected to psychometric testing, anchor validation, and endpoint efficiency comparison to TTFSE. SxBAF exhibited greater efficiency than time to first symptomatic event for treatment difference detection, resulting in smaller projected clinical trial sample sizes. SxBAF correlated well with device-detected AF burden (Spearman's ρ=0.74; P< 0.0001), while permitting gradation in symptom severity. SxBAF also identified treatment group differences in cardiovascular serious adverse events and AF interventions and in recent-onset AF provided specificity for AF- or worsening HF-associated symptoms (P< 0.001 for each). Measurement of symptoms burden throughout clinical trial follow-up is feasible in AF/HF and should be useful for evaluating patient-centered outcomes in AF prevention trials. (Genetically Targeted Therapy for the Prevention of Symptomatic Atrial Fibrillation in Patients With HeartFailure [GENETIC-AF]; NCT01970501).

An evaluation of measurement invariance of Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition borderline personality disorder criteria across heterosexual, lesbian, gay, and bisexual adults.

There is a robust, yet poorly understood relationship between nonheterosexual orientation and borderline personality disorder (BPD), with lesbian, gay, and bisexual individuals evidencing greater BPD symptoms compared to heterosexual individuals. Recent evidence suggests possible psychometric bias in BPD diagnostic criteria leading to greater endorsement among sexual minority individuals, which hinders researchers' ability to make valid group comparisons. The present study utilized an epidemiological sample of 35,995 men and women to evaluate the extent of differential item functioning (DIF) among BPD criteria across sexual orientation groups using a multiple indicators multiple causes approach. All criteria except affective instability and emptiness indicated DIF for at least one sexual minority focal group, although both demonstrated DIF in sensitivity analyses. DIF was most consistently indicated for suicidality, efforts to avoid abandonment, and impulsivity. Contrary to predictions, DIF was mostly nonuniform with greater item discrimination for sexual minority groups compared to their heterosexual counterparts. Finally, all estimated effect sizes were small, suggesting that DIF was not practically meaningful and unlikely to impact the validity of group comparisons for BPD criteria across heterosexual and nonheterosexual men and women. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Investigation on flow–acoustic resonance behaviors inside ducts with tandem cavities using a high-order spectral/hp element method

This study numerically investigates the flow–acoustic resonance behaviors inside ducts with tandem cavities, containing the flow-excited acoustic eigenmodes and elevated flow dynamics under self-sustained acoustic forcing. An advanced high-order spectral/hp element method integrated with implicit large-eddy simulations was utilized to solve the nonlinear compressible Navier–Stokes equations, which effectively identified the fully coupled self-sustained flow–acoustic resonance fields. The benchmark shallow cavity configuration with a length-to-depth (L/D) ratios of 2 was motivated by the experimental findings from Shaaban and Ziada [“Fully developed building unit cavity source for long multiple shallow cavity configurations,” Phys. Fluids 30, 086105 (2018)], in which the intensive flow–acoustic resonance was occurred at a Reynolds number of 1.3×105, and we further investigated three deeper cavity configurations with L/D of 1, 2/3, and 1/2 for numerical validation and further comparison. Subsequently, aeroacoustic characteristics were assessed by analyzing the wall pressure fluctuations, indicating broader resonance regions and augmented pressure pulsation amplitudes extending from main duct to local cavity volumes with larger cavity depths. As feedback, the intensified acoustic forcing can modulate the cavity flow dynamics into stronger fluctuation levels. Furthermore, the spectral proper orthogonal decomposition analysis was conducted on the pressure fields and velocity fields, respectively. The significant fluctuations in acoustic pressure were linked to transitional acoustic modes that were present as global modes in the main duct and local modes in tandem cavities. As for velocity analysis, coherent vortex structures were extracted along the cavity entrances. These vortex structures caused progressively amplified velocity fluctuations and classified the shear layers into two dynamic motions, i.e., a flapping motion in shallow cavities and a rolling-up motion in deep cavities.

Are opioid use disorder assessments in the emergency department biased? An examination across sex, race, and employment status using item response theory

BackgroundIn acute care settings such as the emergency department (ED), the absence of test bias is essential for any assessment. Unbiased assessments of the DSM-5 criteria for opioid use disorder (OUD) are crucial to ensure valid comparison of OUD severity, though focused tests of potential bias are lacking. MethodThis study used item response theory (IRT) to examine potential test bias across patient characteristics (i.e., sex, race, employment status) in a sample of 518 individuals aged 18+ who reported opioid use within the prior 12 months. Participants indicated the presence or absence of the 11 DSM-5 OUD criteria, as well as self-reported demographic information. ResultsIRT indicated a lack of test bias (i.e., measurement invariance) for DSM-5 OUD across biological sex, race, and employment status. The lack of test bias permitted the groups to be compared on OUD severity across demographic groups. Males, White individuals, and unemployed individuals reported higher OUD severity. ConclusionsFindings indicate the DSM-5 OUD criteria are unbiased across sex, race, and employment status. Thus, these criteria can be used in the ED to compare differences in OUD severity across key demographic groups. The current study provides empirical support for using OUD assessments in ED settings to inform in-vivo clinical decisions across different sexes, races, and employment statuses. Future studies are needed to corroborate this finding in additional samples and across additional demographic characteristics (e.g., ethnicity, age).

Critical Assessment on the Stability and Convergence of the Conventional Gear Tooth Contact Analysis

Mathematical modeling of gear engagement is crucial during design to ensure optimal performance in manufacturing. This study reproduces the conventional tooth contact analysis (TCA) model, highlighting convergence issues in parallel-axis gears and limitations in local synthesis methods. The research critically analyzes the TCA method, which solves five nonlinear equations to assess performance and accuracy. Simulations replicate the conditions of previous studies to ensure valid comparisons. Initial guess values are randomly generated within a specific range to guide the iterative process toward convergence, with this range progressively narrowed to improve computational efficiency and accuracy. Results indicate that the TCA approach is highly sensitive to initial guess values, particularly the starting angular position. Convergence issues arise from the complexity of nonlinear equations and multiple roots. This can lead to divergence or reverting to the initial guess when values deviate significantly from the true solution.

Cross-institutional evaluation of deep learning and radiomics models in predicting microvascular invasion in hepatocellular carcinoma: validity, robustness, and ultrasound modality efficacy comparison

PurposeTo conduct a head-to-head comparison between deep learning (DL) and radiomics models across institutions for predicting microvascular invasion (MVI) in hepatocellular carcinoma (HCC) and to investigate the model robustness and generalizability through rigorous internal and external validation.MethodsThis retrospective study included 2304 preoperative images of 576 HCC lesions from two centers, with MVI status determined by postoperative histopathology. We developed DL and radiomics models for predicting the presence of MVI using B-mode ultrasound, contrast-enhanced ultrasound (CEUS) at the arterial, portal, and delayed phases, and a combined modality (B + CEUS). For radiomics, we constructed models with enlarged vs. original regions of interest (ROIs). A cross-validation approach was performed by training models on one center’s dataset and validating the other, and vice versa. This allowed assessment of the validity of different ultrasound modalities and the cross-center robustness of the models. The optimal model combined with alpha-fetoprotein (AFP) was also validated. The head-to-head comparison was based on the area under the receiver operating characteristic curve (AUC).ResultsThirteen DL models and 25 radiomics models using different ultrasound modalities were constructed and compared. B + CEUS was the optimal modality for both DL and radiomics models. The DL model achieved AUCs of 0.802–0.818 internally and 0.667–0.688 externally across the two centers, whereas radiomics achieved AUCs of 0.749–0.869 internally and 0.646–0.697 externally. The radiomics models showed overall improvement with enlarged ROIs (P < 0.05 for both CEUS and B + CEUS modalities). The DL models showed good cross-institutional robustness (P > 0.05 for all modalities, 1.6–2.1% differences in AUC for the optimal modality), whereas the radiomics models had relatively limited robustness across the two centers (12% drop-off in AUC for the optimal modality). Adding AFP improved the DL models (P < 0.05 externally) and well maintained the robustness, but did not benefit the radiomics model (P > 0.05).ConclusionCross-institutional validation indicated that DL demonstrated better robustness than radiomics for preoperative MVI prediction in patients with HCC, representing a promising solution to non-standardized ultrasound examination procedures.

A new validated TRNSYS module for phase change material-filled multi-glazed windows

Incorporating phase change materials (PCM) into multi-glazed windows (MW) has been recognized as an effective way to regulate the indoor thermal environment and reduce the energy consumption of buildings. However, the design and calculation of PCM-filled multi-glazed windows (MW + PCM) remain challenging due to the lack of general transient simulation methods and integrated tools. To address this issue, the present study developed a dynamic coupled thermal and optical model for MW + PCM based on the enthalpy-temperature and interface energy balance methods. Then, the model was implemented in C++ and integrated into the TRNSYS software as a new module (Type 2602). In parallel, a test platform comprising two test chambers was established to evaluate the effect of PCM on the thermal and optical properties of the window and to validate the newly proposed module. It was found that compared to the traditional triple-glazed window (TW), the PCM-filled triple-glazed window (TW + PCM) effectively reduces the irradiance through the window by 34.17 % and the maximum temperature of the inner surface by approximately 5.56 °C. While the increase in external surface radiation flux density from 600 to 1200 W/m2 has a diminishing effect on the heat flux variation rate during the heat dissipation process for TW + PCM, it effectively shortens the latent heat absorption time by 0.50 h and increases the temperature difference between the inner and outer surfaces by 3.02 °C. The new module demonstrates satisfactory accuracy through experimental validation and simulation comparisons, with the maximum heat flux coefficient of variance of the root mean square error (CV(RMSE)) less than 6.35 % and 11.25 % under TW and TW + PCM configurations.

Exploring the Learning Curve of Dental Implant Placement Using a Task-Autonomous Robotic System Among Young Dentists From Different Specialties-A Pilot Module Study.

The learning curve effect of dynamic computer-assisted implant surgery (D-CAIS) was observed among inexperienced novice surgeons. The learning curves can provide valuable information for novice surgeons and valid comparisons between new and conventional techniques. Recently, robotic computer-assisted implant surgery (R-CAIS) has shown promise as a novel dental implant surgical technique for both partially and edentulous patients. However, its learning curve remains unknown. The aim of this study was to explore the learning curve of dental implant placement surgery with a task-autonomous robotic system among young dentists with different specialties. Four young dentists (mean age: 25.3 ± 1.5 years at the beginning of their first attempt) with equal representation of males and females and with different specialties participated in this study. None of the participants had prior experience in R-CAIS. Each operator placed eight implants over eight attempts using a semi-active task-autonomous robotic system. Among the eight implants, four were straight lateral incisor implants, and four were 30°-tilted premolar implants. The implants were placed in each dental quadrant of the maxillary and mandibular jaw modules. The operation time was recorded. Coronal, apical, and angular deviations between the planned and actual sites of implant placement were measured by merging preoperative and postoperative cone-beam computed tomography (CBCT) scans. The data were analyzed with repeated-measures ANOVA (α = 0.05). The mean time for implant placement was associated with the number of attempts (p < 0.01). The time taken for the second attempt was significantly shorter than that of the first attempt (33.26 vs. 30.47 min; p < 0.001) then it plateaued. Three-dimensional (3D) angular (p = 0.31), coronal deviation (p = 0.26), and apical deviation (p = 0.06) did not differ significantly among attempts. The mean values and standard deviations of 3D coronal deviation, 3D apical deviation, and 3D angular deviation were 0.71 ± 0.31 mm, 0.72 ± 0.30 mm, and 0.94 ± 0.58°, respectively. Neither the position of the jaw (p > 0.59) nor the tilt angle of the implant (straight or 30°-tilted, p > 0.85) was related to implant placement accuracy. Dentists quickly learned the basic workflow of R-CAIS and thus facilitated the clinicians in the mastery of implant placement on edentulous jaw modules, leading to a comparable operating speed and high precision. Moreover, the accuracy of placement of straight and tilted implants in both the maxilla and mandible with R-CAIS was satisfactory.