Internal Data Research Articles

TMTV-Net: fully automated total metabolic tumor volume segmentation in lymphoma PET/CT images - a multi-center generalizability analysis.

Total metabolic tumor volume (TMTV) segmentation has significant value enabling quantitative imaging biomarkers for lymphoma management. In this work, we tackle the challenging task of automated tumor delineation in lymphoma from PET/CT scans using a cascaded approach. Our study included 1418 2-[18F]FDG PET/CT scans from four different centers. The dataset was dividedinto 900 scans for development/validation/testing phasesand 518 for multi-center external testing. The former consisted of 450 lymphoma, lung cancer, and melanoma scans, along with 450 negative scans, while the latter consisted of lymphoma patients from different centers with diffuse large B cell, primary mediastinal large B cell, and classic Hodgkin lymphoma cases. Our approach involves resampling PET/CT images into different voxel sizes in the first step, followed by training multi-resolution 3D U-Nets on each resampled dataset using a fivefold cross-validation scheme. The models trained on different data splits were ensemble. After applying soft voting to the predicted masks, in the second step, we input the probability-averaged predictions, along with the input imaging data, into another 3D U-Net. Models were trained with semi-supervised loss. We additionally considered the effectiveness of using test time augmentation (TTA) to improve the segmentation performance after training. In addition to quantitative analysis including Dice score (DSC) and TMTV comparisons, the qualitative evaluation was also conducted by nuclear medicine physicians. Our cascaded soft-voting guided approach resulted in performance with an average DSC of 0.68 ± 0.12 for the internal test data from developmental dataset, and an average DSC of 0.66 ± 0.18 on the multi-site external data (n = 518), significantly outperforming (p < 0.001) state-of-the-art (SOTA) approaches including nnU-Net and SWIN UNETR. While TTA yielded enhanced performance gains for some of the comparator methods, its impact on our cascaded approach was found to be negligible (DSC: 0.66 ± 0.16). Our approach reliably quantified TMTV, with a correlation of 0.89 with the ground truth (p < 0.001). Furthermore, in terms of visual assessment, concordance between quantitative evaluations and clinician feedback was observed in the majority of cases. The average relative error (ARE) and the absolute error (AE) in TMTV prediction on external multi-centric dataset were ARE = 0.43 ± 0.54 and AE = 157.32 ± 378.12 (mL) for all the external test data (n = 518), and ARE = 0.30 ± 0.22 and AE = 82.05 ± 99.78 (mL) when the 10% outliers (n = 53) were excluded. TMTV-Net demonstrates strong performance and generalizability in TMTV segmentation across multi-site external datasets, encompassing various lymphoma subtypes. A negligible reduction of 2% in overall performance during testing on external data highlights robust model generalizability across different centers and cancer types, likely attributable to its training with resampled inputs. Our model is publicly available, allowing easy multi-site evaluation and generalizability analysis on datasets from different institutions.

An External Validation Study for Automated Segmentation of Vestibular Schwannoma.

To validate how an automated model for vestibular schwannoma (VS) segmentation developed on an external homogeneous dataset performs when applied to internal heterogeneous data. The external dataset comprised 242 patients with previously untreated, sporadic unilateral VS undergoing Gamma Knife radiosurgery, with homogeneous magnetic resonance imaging (MRI) scans. The internal dataset comprised 10 patients from our institution, with heterogeneous MRI scans. An automated VS segmentation model was developed on the external dataset. The model was tested on the internal dataset. Dice score, which measures agreement between ground truth and predicted segmentations. When applied to the internal patient scans, the automated model achieved a mean Dice score of 61% across all 10 images. There were three tumors that were not detected. These tumors were 0.01 ml on average (SD = 0.00 ml). The mean Dice score for the seven tumors that were detected was 87% (SD = 14%). There was one outlier with Dice of 55%-on further review of this scan, it was discovered that hyperintense petrous bone had been included in the tumor segmentation. We show that an automated segmentation model developed using a restrictive set of siloed institutional data can be successfully adapted for data from different imaging systems and patient populations. This is an important step toward the validation of automated VS segmentation. However, there are significant shortcomings that likely reflect limitations of the data used to train the model. Further validation is needed to make automated segmentation for VS generalizable.

Financial impact of medication reviews by clinical pharmacists to reduce in-hospital adverse drug events: a return-on-investment analysis

BackgroundAdverse drug events contribute to rising health care costs. Clinical pharmacists can reduce their risks by identifying and solving drug-related problems (DRPs) through medication review.AimTo develop an economic model to determine whether medication reviews performed by clinical pharmacists could lead to a reduction in health care costs associated with the prevention of potential adverse drug events.MethodTwo pharmacists performed medication reviews during ward rounds in an internal medicine setting over one year. Avoided costs were estimated by monetizing five categories of DRPs (improper drug selection, drug interactions, untreated indications, inadequate dosages, and drug use without an indication). An expert panel assessed potential adverse drug events and their probabilities of occurrence for 20 randomly selected DRPs in each category. The costs of adverse drug events were extracted from internal hospital financial data. A partial economic study from a hospital perspective then estimated the annual costs avoided by resolving DRPs identified by 3 part-time clinical pharmacists (0.9 full-time equivalent) from 2019 to 2020. The return on investment (ROI) of medication review was calculated.ResultsThe estimated annual avoided costs associated with the potential adverse drug events induced by 676 DRPs detected was € 304,170. The cost of a 0.9 full-time equivalent clinical pharmacist was € 112,408. Extrapolated to 1 full-time equivalent, the annual net savings was € 213,069 or an ROI of 1–1.71. Sensitivity analyses showed that the economic model was robust.ConclusionThis economic model revealed the positive financial impact and favorable return on investment of a medication review intervention performed by clinical pharmacists. These findings should encourage the future deployment of a pharmacist-led adverse drug events prevention program.

[Construction and verification of a nomogram of factors influencing the risk of death in patient with sepsis-associated thrombocytopenia].

To construct a nomogram prediction model for predicting the risk of death in patients with sepsis-associated thrombocytopenia (SAT) in intensive care unit (ICU) for early indentification and active intervention. Clinical data of SAT patients admitted to ICU of the First Affiliated Hospital of Nanjing Medical University from December 2019 to August 2021 were retrospectively collected, including demographic data, laboratory indicators, etc. According to the prognosis at 28 days, the patients were divided into the death group and the survival group, and the differences of clinical variables between the two groups were compared. Multivariate Logistic regression analysis was performed to analyze the independent risk factors influencing mortality of patients within 28 days, then a nomogram predictive model was constructed and its performance was verified with internal data. Receiver operator characteristic curve (ROC curve) was used to evaluate the diagnostic effectiveness of the nomogram model, and the clinical applicability of this model was evaluated by clinical decision curve analysis (DCA). A total of 275 patients were included, with 95 deaths at 28 days and a 28-day mortality of 34.5%. Compared with the survival group, acute physiology and chronic health evaluation II (APACHE II), sequential organ failure assessment (SOFA), lactic acid (Lac), platelet distribution width (PDW) on day 5 of ICU admission, blood urea nitrogen (BUN), total bilirubin (TBIL), aspartate aminotransferase (AST), C-reactive protein (CRP) of patients in the death group were higher, activated partial thromboplastin time (APTT) and prothrombin time (PT) were longer, platelet count (PLT) on day 3 and day 5 of ICU admission, direct bilirubin (DBIL), fibrinogen (FIB) were lower, the history of chronic lung disease, mixed site infection, lung infection, bloodstream infection, Gram-negative bacterial infection and fungal infection accounted for a higher proportion, the history of diabetes mellitus, urinary tract infection and no pathogenic microorganisms cultured accounted for a lower proportion, and the proportion of the use of vasoactive drugs, mechanical ventilation (MV), continuous renal replacement therapy (CRRT), bleeding events and platelet transfusion were higher. Multivariate Logistic regression analysis showed that APACHE II score at the day of ICU admission [odds ratio (OR) = 1.417, 95% confidence interval (95%CI) was 1.153-1.743, P = 0.001], chronic lung disease (OR = 72.271, 95%CI was 4.475-1 167.126, P = 0.003), PLT on day 5 of ICU admission (OR = 0.954, 95%CI was 0.922-0.987, P = 0.007), vasoactive drug (OR = 622.943, 95%CI was 10.060-38 575.340, P = 0.002), MV (OR = 91.818, 95%CI was 3.973-2 121.966, P = 0.005) were independent risk factors of mortality in SAT patients. The above independent risk factors were used to build a nomogram prediction model, and the area under the curve (AUC), sensitivity and specificity were 0.979, 94.7% and 91.7%, respectively, suggesting that the model had good discrimination. The Hosmer-Lemeshow goodness of fit test showed a good calibration with P > 0.05. At the same time, DCA showed that the nomogram model had good clinical applicability. Patients with SAT has a higher risk of death. The nomogram model based on APACHE II score at the day of ICU admission, chronic lung disease, PLT on day 5 of ICU admission, the use of vasoactive drug and MV has good clinical significance for the prediction of 28-day mortality, and the discrimination and calibration are good, however, further verification is needed.

Leadership During ERM Transition Implementation in The Public Sector

This research aims to determine the appropriate leadership style in the public sector transition period from implementing vernacular risk management to formal risk management or ERM. This leadership style also considers leadership factors, and implementing a risk culture has become the main factor in implementing ERM in various studies. The research method uses a descriptive qualitative approach, with a research strategy in the form of case studies. The data used is primary data and secondary data. Primary data is the results of interviews with 16 respondents who are parties who participate in managing risks at the ZYX ministry. Secondary data is all internal and external data related to MRPN. The general conclusion is that Agile leadership is needed so that ZYX Ministry can adapt to the VUCA era. Leaders with agility can guide their organizations to become faster, more adaptive, and more productive. Agile Governance is the key to responsive and proactive governance.

Cost-Effective Acquisition of First-Party Data for Business Analytics

Customer data acquisition is an important task in data-driven business analytics. Recently, there has been a growing interest in the effective use of an organization’s internal customer data, also known as first-party data. This work studies the acquisition of new data for business analytics based on first-party data resource. We address issues related to both acquisition cost and data quality. To reduce acquisition cost, we consider using auction-based methods, such as the generalized second price (GSP) auction, for acquiring data with differential prices for different customers. We find that the GSP-based data acquisition method incurs a lower cost and/or achieves a higher response rate than fixed price methods. To maximize data quality, we propose novel optimization models for different data acquisition methods and data quality measures. The proposed models maximize the quality of the acquired data while satisfying budget constraints. We derive and discuss the solutions to the optimization models analytically and provide managerial insights from the solutions. The proposed approach is effective in increasing customer responses, reducing selection bias, and enabling more accurate estimation and prediction for business analytics. The results of the experimental evaluation demonstrate the advantage of the proposed approach over existing data acquisition methods. History: Accepted by Ram Ramesh, Area Editor for Data Science and Machine Learning. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.0037 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2022.0037 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Disulfidptosis-related genes of prognostic signature and immune infiltration features in hepatocellular carcinoma supported by bulk and single-cell RNA sequencing data.

Disulfidptosis is a new type of cellular death triggered in response to disulfide stress and is strongly linked to the progression of malignancies. Hepatocellular carcinoma (HCC) is a very common malignancy. Some reports have suggested a link between disulfidptosis-related genes (DRGs) and cancer; however, further research needs to be conducted. In this study, HCC data from the Cancer Genome Atlas-Liver Hepatocellular Carcinoma and Gene Expression Omnibus data sets were collected and analyzed. A univariate Cox regression analysis, least absolute shrinkage and selection operator, and multivariate Cox regression analysis were conducted to identify the hub DRGs signature for prognosis. The HCC patients were allocated to high- and low-risk groups based on their disulfidptosis risk scores. The model was validated with a high degree of precision using both internal and external validation data sets. "ESTIMATE" and "CIBERSORT" packages were employed to assess the immunological landscapes and immune cell infiltration. The IMvigor210 cohort was chosen to validate the immunotherapy results. A drug sensitivity analysis was conducted to identify targeted medications. The expression of the hub DRGs in the HCC cells was confirmed using cytological techniques. The bioinformatic analysis revealed that 16 genes showed differential expression. A prognostic model was developed based on four genes: RPN1, SLC2A1, SLC2A4, and SLC7A11. A notable difference in prognosis was observed between the two risk groups. Based on the results of the immune microenvironment, tumor mutation burden, immunotherapy, and drug screening analyses, the DRGs signature can be employed in HCC immunotherapy decision making. Further, the expression levels of the hub DRGs were significantly upregulated in the HCC cells. Our four-DRGs signature could be used to predict HCC prognosis. Further, this study showed that the hub DRGs could serve as biomarkers for immunotherapy prediction and could potentially guide targeted therapies.

The value of diffusion-weighted imaging in the natural history of meningiomas: a predictor of tumor growth.

The general trend in meningioma treatment is shifting from surgery to active surveillance. However, the natural history of meningioma still needs to be clarified, and a simple, practical method is needed to identify fast-growing tumors. The authors aimed to determine whether diffusion-weighted imaging (DWI) could be a valuable imaging modality for predicting meningioma growth. Consecutive asymptomatic patients with a meningioma diagnosed on MRI and followed up at the authors' institution between July 2011 and July 2019 were eligible for inclusion in this retrospective study. Univariable and multivariable Cox regression analyses were used to explore whether the relative apparent diffusion coefficient (rADC) was an independent predictor of meningioma growth. Correlations between tumor growth rate (TGR), tumor volume doubling time (VDT), Ki-67, and rADC were assessed using the Pearson correlation coefficient. The predictive ability of rADC was evaluated using receiver operating characteristic (ROC) curves and validated with internal validation data. Sixty-four patients (47 females, 17 males) with a mean age of 62.2 ± 1.4 years were included in this study. Univariable and multivariable analyses revealed that rADC was an independent predictor of meningioma growth (p < 0.05). ROC curve analysis showed that baseline rADC had good predictive power for growing meningiomas (AUC = 0.88, 95% CI 0.78-0.96), as well as slow- or fast-growing meningiomas (AUC = 0.83, 95% CI 0.59-0.98). Moreover, rADC still had a good ability to discriminate between growing and nongrowing meningiomas in the validation set (AUC = 0.85, 95% CI 0.64-1.00). In the 20 patients with tumor growth, baseline rADC was moderately negatively correlated with TGR (r = -0.50, p = 0.02) and strongly positively correlated with VDT (r = 0.63, p = 0.003). Moreover, Ki-67 was significantly associated with rADC in 8 patients who had undergone surgery (r = -0.75, p = 0.03). In asymptomatic meningiomas, the lower the rADC at baseline, the faster the TGR and the shorter the VDT. DWI could be a valuable tool in predicting meningioma growth in asymptomatic patients.

An enhanced state-aware model learning approach for security analysis in lightweight protocol implementations

Owing to the emergence and rapid advances of new-generation information and digitalization technologies, the concept of model-driven digital twin has received widespread attentions and is developing vigorously. Driven by data and simulators, the digital twin can create the virtual twins of physical objects to perform monitoring, simulation, prediction, optimization, and so on. Hence, the application of digital twin can increase efficiency and security of systems by providing reliable model and decision supports. In this paper, we propose a state-aware model learning method to simulate and analyze the lightweight protocol implementations in edge/cloud environments. We introduce the data flow of program execution and network interaction inputs/outputs (I/O) into the extended finite state machine (EFSM) to expand the modeling scope and insight. We aim to calibrate the states and construct an accurate state-machine model using a digital twin based layered approach to reasonably reflect the correlation of a device’s external behavior and internal data. This, in turn, improves our ability to verify the logic and evaluate the security for protocol implementations. This method firstly involves instrumenting the target device to monitor variable activity during its execution. We then employ learning algorithms to produce multiple rounds of message queries. Both the I/O data corresponding to these query sequences and the state calibration information derived from filtered memory variables are obtained through the mapper and execution monitor, respectively. These two aspects of information are combined to dynamically and incrementally construct the protocol’s state machine. We apply this method to develop SALearn and evaluate the effectiveness of SALearn on two lightweight protocol implementations. Our experimental results indicate that SALearn outperforms existing protocol model learning tools, achieving higher learning efficiency and uncovering more interesting states and security issues. In total, we identified two violation scenarios of rekey logic. These situations also reflect the differences in details between different implementations.

Borrowing networks for innovation: The role of attention allocation in secondhand brokerage

AbstractResearch SummaryThis study investigates how people connecting with brokers who span structural holes in a firm's communication networks can enhance their innovative performance. Through sociometric data from a large software company, we observe that individuals who pay attention to information from brokers achieve higher innovative performance compared with those who ignore such information. The advantage of paying attention to brokers' information is more pronounced when people operate in highly constrained networks. Yet, our post hoc analysis reveals that people typically allocate less attention to information from brokers than from local colleagues—they systematically do the opposite of what they should do for innovative performance. Our findings regarding the role of attention for secondhand brokerage make significant contributions to studies of networks and innovation for behavioral strategy.Managerial SummaryThis study investigates how innovation is influenced by paying attention to brokers within a company. Brokers are individuals who bridge different groups, providing access to varied information. Analysis of internal data from a large software company reveals that employees who heed the ideas of brokers tend to be more innovative, especially in close‐knit networks. However, there is a notable tendency for people to ignore brokers, favoring information from familiar colleagues instead. This indicates a significant attention bias, where the potentially most valuable sources for innovation are often unattended. Our research highlights the critical need for a strategic approach to distributing attention within company networks, which can significantly boost innovative performance.

Research on Government Data Sharing Mechanism Based on Smart Contracts

Exploring mechanisms for internal data sharing within government departments is important in advancing digital and intelligent society. This paper is based on the establishment of decentralized nodes on the external network of government departments, constructing a decentralized node, and establishing a government internal data sharing model based on blockchain. Subsequently, integrating attribute fields from government data into the shared model, accompanied by the formulation of data-sharing rules through smart contracts, serves to streamline the implementation of efficient and secure cross-validation mechanisms across diverse departments. Finally, this article concludes by conducting a model performance testing experiment, evaluating the model from three perspectives: storage cost, blockchain performance, and security analysis. The test results show that our model enhances the efficiency of querying and retrieving data within the government's internal data-sharing system, effectively addressing challenges such as low efficiency, high costs, and issues related to the security and real-time aspects of data sharing within the government. Overall, our article provides a new way of thinking about government data sharing.

Changing interim monitoring in response to internal clinical trial data.

Designing clinical trials for emerging infectious diseases such as COVID-19 is challenging because information needed for proper planning may be lacking. Pre-specified adaptive designs can be attractive options, but what happens if a trial with no such design needs to be modified? For example, unexpectedly high efficacy (approximately 95%) in two COVID-19 vaccine trials might cause investigators in other COVID-19 vaccine trials to increase the number of interim analyses to allow earlier stopping for efficacy. If such a decision is based solely on external data, there are no issues, but what if internal trial data by arm are also examined? Fortunately, the conditional error principle of Müller and Schäfer (2004) can be used to ensure no inflation of the type 1 error rate, even if no interim analyses were planned. We study the properties, including limitations, of this method. We provide a shiny app to evaluate changes in timing of interim analyses in response to outcome data by arm in clinical trials.

Robust data integration from multiple external sources for generalized linear models with binary outcomes.

We aim to estimate parameters in a generalized linear model (GLM) for a binary outcome when, in addition to the raw data from the internal study, more than 1 external study provides summary information in the form of parameter estimates from fitting GLMs with varying subsets of the internal study covariates. We propose an adaptive penalization method that exploits the external summary information and gains efficiency for estimation, and that is both robust and computationally efficient. The robust property comes from exploiting the relationship between parameters of a GLM and parameters of a GLM with omitted covariates and from downweighting external summary information that is less compatible with the internal data through a penalization. The computational burden associated with searching for the optimal tuning parameter for the penalization is reduced by using adaptive weights and by using an information criterion when searching for the optimal tuning parameter. Simulation studies show that the proposed estimator is robust against various types of population distribution heterogeneity and also gains efficiency compared to direct maximum likelihood estimation. The method is applied to improve a logistic regression model that predicts high-grade prostate cancer making use of parameter estimates from 2 external models.

Development of a generative deep learning model to improve epiretinal membrane detection in fundus photography

BackgroundThe epiretinal membrane (ERM) is a common retinal disorder characterized by abnormal fibrocellular tissue at the vitreomacular interface. Most patients with ERM are asymptomatic at early stages. Therefore, screening for ERM will become increasingly important. Despite the high prevalence of ERM, few deep learning studies have investigated ERM detection in the color fundus photography (CFP) domain. In this study, we built a generative model to enhance ERM detection performance in the CFP.MethodsThis deep learning study retrospectively collected 302 ERM and 1,250 healthy CFP data points from a healthcare center. The generative model using StyleGAN2 was trained using single-center data. EfficientNetB0 with StyleGAN2-based augmentation was validated using independent internal single-center data and external datasets. We randomly assigned healthcare center data to the development (80%) and internal validation (20%) datasets. Data from two publicly accessible sources were used as external validation datasets.ResultsStyleGAN2 facilitated realistic CFP synthesis with the characteristic cellophane reflex features of the ERM. The proposed method with StyleGAN2-based augmentation outperformed the typical transfer learning without a generative adversarial network. The proposed model achieved an area under the receiver operating characteristic (AUC) curve of 0.926 for internal validation. AUCs of 0.951 and 0.914 were obtained for the two external validation datasets. Compared with the deep learning model without augmentation, StyleGAN2-based augmentation improved the detection performance and contributed to the focus on the location of the ERM.ConclusionsWe proposed an ERM detection model by synthesizing realistic CFP images with the pathological features of ERM through generative deep learning. We believe that our deep learning framework will help achieve a more accurate detection of ERM in a limited data setting.

A Novel Hybrid Deep Learning Method for Predicting the Flow Fields of Biomimetic Flapping Wings.

The physics governing the fluid dynamics of bio-inspired flapping wings is effectively characterized by partial differential equations (PDEs). Nevertheless, the process of discretizing these equations at spatiotemporal scales is notably time consuming and resource intensive. Traditional PDE-based computations are constrained in their applicability, which is mainly due to the presence of numerous shape parameters and intricate flow patterns associated with bionic flapping wings. Consequently, there is a significant demand for a rapid and accurate solution to nonlinear PDEs, to facilitate the analysis of bionic flapping structures. Deep learning, especially physics-informed deep learning (PINN), offers an alternative due to its great nonlinear curve-fitting capability. In the present work, a hybrid coarse-data-driven physics-informed neural network model (HCDD-PINN) is proposed to improve the accuracy and reliability of predicting the time evolution of nonlinear PDEs solutions, by using an order-of-magnitude-coarser grid than traditional computational fluid dynamics (CFDs) require as internal training data. The architecture is devised to enforce the initial and boundary conditions, and incorporate the governing equations and the low-resolution spatiotemporal internal data into the loss function of the neural network, to drive the training. Compared to the original PINN with no internal data, the training and predicting dynamics of HCDD-PINN with different resolutions of coarse internal data are analyzed on the problem relevant to the two-dimensional unsteady flapping wing, which involves unsteady flow features and moving boundaries. Additionally, a hyper-parametrical study is conducted to obtain an optimal model for the problem under consideration, which is then utilized for investigating the effects of the snapshot and fraction of the coarse internal data on the HCDD-PINN's performances. The results show that the proposed framework has a sufficient stability and accuracy for solving the considered biomimetic flapping-wing problem, and its great potential means that it can be considered as an alternative to accelerate or replace traditional CFD solvers in the future. The interested variables of the flow field at any instant can be rapidly obtained by the trained HCDD-PINN model, which is superior to the traditional CFD method that usually needs to be re-run. For the three-dimensional and optimization problems of flapping wings, the advantages of the proposed method are supposedly even more apparent.

Revisión de literatura sobre metodologías de diseño de user personas para museos y bibliotecas

The persona design technique, coming from the user-centered design environment, can be used as a tool in the design of improvements for the heritage dissemination, cultural communication and marketing of information spaces such as museums and libraries (user personas, visitor personas, patron personas). The components of personas designed for these contexts are presented, and a systematic review of literature is carried out to detect what different methodologies (qualitative, quantitative or with mix-methods) can be used to construct personas, based on internal or external data, within this cultural context, in 31 scientific publications.

Anisotropic span embeddings and the negative impact of higher-order inference for coreference resolution: An empirical analysis

Abstract Coreference resolution is the task of identifying and clustering mentions that refer to the same entity in a document. Based on state-of-the-art deep learning approaches, end-to-end coreference resolution considers all spans as candidate mentions and tackles mention detection and coreference resolution simultaneously. Recently, researchers have attempted to incorporate document-level context using higher-order inference (HOI) to improve end-to-end coreference resolution. However, HOI methods have been shown to have marginal or even negative impact on coreference resolution. In this paper, we reveal the reasons for the negative impact of HOI coreference resolution. Contextualized representations (e.g., those produced by BERT) for building span embeddings have been shown to be highly anisotropic. We show that HOI actually increases and thus worsens the anisotropy of span embeddings and makes it difficult to distinguish between related but distinct entities (e.g., pilots and flight attendants). Instead of using HOI, we propose two methods, Less-Anisotropic Internal Representations (LAIR) and Data Augmentation with Document Synthesis and Mention Swap (DSMS), to learn less-anisotropic span embeddings for coreference resolution. LAIR uses a linear aggregation of the first layer and the topmost layer of contextualized embeddings. DSMS generates more diversified examples of related but distinct entities by synthesizing documents and by mention swapping. Our experiments show that less-anisotropic span embeddings improve the performance significantly (+2.8 F1 gain on the OntoNotes benchmark) reaching new state-of-the-art performance on the GAP dataset.

The application of impantable sensors in the musculoskeletal system: a review.

As the population ages and the incidence of traumatic events rises, there is a growing trend toward the implantation of devices to replace damaged or degenerated tissues in the body. In orthopedic applications, some implants are equipped with sensors to measure internal data and monitor the status of the implant. In recent years, several multi-functional implants have been developed that the clinician can externally control using a smart device. Experts anticipate that these versatile implants could pave the way for the next-generation of technological advancements. This paper provides an introduction to implantable sensors and is structured into three parts. The first section categorizes existing implantable sensors based on their working principles and provides detailed illustrations with examples. The second section introduces the most common materials used in implantable sensors, divided into rigid and flexible materials according to their properties. The third section is the focal point of this article, with implantable orthopedic sensors being classified as joint, spine, or fracture, based on different practical scenarios. The aim of this review is to introduce various implantable orthopedic sensors, compare their different characteristics, and outline the future direction of their development and application.

GMOCU: Digital Documentation, Management, and Biological Risk Assessment of Genetic Parts.

The continuous evolution of molecular biology and gene synthesis methods paired with an ever-increasing potential of synthetic biology approaches and genome engineering toolkits enables the rapid design of genetic bioparts and genetically modified organisms. Although various software solutions assist with specific design tasks and challenges, lab internal documentation and ensuring compliance with governmental regulations on biosafety assessment of the generated organisms remain the responsibility of individual academic researchers. This results in inconsistent and redundant documentation regimes and a significant time and labor burden. GMOCU (GMO documentation) is a standardized semi-automatic user-oriented software approach -written in Python and freely available- that unifies lab internal data documentation on genetic parts and genetically modified organisms (GMOs). It automatizes biological risk evaluations and maintains a shared up-to-date inventory of bioparts for team-wide data navigation and sharing. GMOCU further enables data export into customizable formats suitable for scientific publications, official biosafety documents, and the research community.

Results from a phase 2 study of triplet blockade of the IL-27, PD-(L)1, and VEGF pathways with casdozokitug (casdozo, SRF388) in combination with atezolizumab (atezo) and bevacizumab (bev) in patients with unresectable, locally advanced, or metastatic hepatocellular carcinoma (uHCC).

470 Background: Casdozo is the first in class and only clinical-stage IL-27 targeting antibody, which neutralizes IL-27 in the tumor microenvironment and stimulates antitumor response. A phase (Ph) 1 study demonstrated a favorable safety profile and antitumor activity alone and in combination with PD-1 blockade in indications known to have high levels of IL-27 pathway activation, such as HCC (NCT04374877). Casdozo induced increases in serum IFN-g and NK cell gene activation, indicating an immune response and reversal of IL-27-mediated immune suppression. IHC evaluation of HCC commercial tissue microarrays revealed that most HCC samples express the target: IL-27+ tumor-associated macrophages (TAMs, internal data). We present the safety and antitumor activity of this immunoregulatory cytokine antagonist given in combination with atezo/bev in uHCC in the open-label phase of a randomized Ph 2 trial. Methods: Patients (pts) with untreated uHCC received casdozo 10 mg/kg, atezo 1200 mg, and bev 15mg/kg IV every 3 weeks. Primary endpoints were safety and tolerability, with key secondary endpoints being investigator-assessed ORR by RECIST1.1 and mRECIST. Results: As of August 9, 2023, 30 pts were treated with casdozo/atezo/bev. Most pts were male (77%), Asian (67%), ECOG PS 1 (77%), had viral etiology (53% HBV, 17% HCV), and had poor-risk disease as evidenced by metastatic spread (70%) and macrovascular invasion (23%). 47% had AFP ≥400ng/mL. Median time on therapy was 28 wks (3-69). Most treatment-related adverse events (TRAEs) were low grade, with proteinuria (33%), fatigue (20%), decreased appetite (20%), and hypertension (20%) being most common. Grade 3/4 TRAEs occurred in 33%, with only hypertension (13%) arising in ≥10% of pts. TRAEs resulting in any study drug discontinuation was 20%. No deaths were attributed to study drugs. In the 29 response-evaluable pts, 67% experienced any tumor shrinkage with a 38% ORR (3 CRs, 5 confirmed PRs, 3 unconfirmed PRs). In the 28 pts evaluable for mRECIST, ORR was 43% (n=12). More than 50% of pts remain on therapy. Available archival tissue of responders (n=4) all expressed IL-27+ TAMs by IHC, including tumors of viral (n=3) and nonviral etiologies (n=1). Conclusions: Triplet blockade of the IL-27, VEGF, and PD-(L)1 pathways with casdozo/atezo/bev has an acceptable safety profile to date with promising antitumor activity in uHCC that warrants continued exploration. Toxicity was consistent with the known profiles of the agents, with no new safety signals identified. Biomarker analysis to evaluate immune response and associations of IL-27 pathway and clinical outcomes is ongoing. Additional clinical studies of casdozo in combination with the PD-1 inhibitor toripalimab are planned. Clinical trial information: NCT05359861 .