Retention Of Data Research Articles

Re-processing Datawell Directional Waverider MkIII heave data affected by breaking waves

Observations of nearshore waves are essential for coastal management, planning and operations. However, shallow water deployments often encounter breaking waves, causing anomalous measurements that are rejected during quality control. A new methodology for re-processing wave data containing breaking waves has been developed, utilising data from the National Network of Regional Coastal Monitoring Programmes of England which operates 37 Datawell Directional Waverider MkIII buoys. Heave data affected by breaking waves is reviewed, anomalous heaves removed, and the remaining data re-processed. Validation through a two-stage process confirms that re-processed wave parameters closely match those derived from the established Datawell processing and ensures strong agreement between parameters derived from shorter re-processed records and full-length data. Applying this new approach to all measured storm wave data, 7.3% (928.5 hours) of storm wave data is recovered across the 37 sites. At some sites, the retention of additional storm wave data has considerable impacts on significant wave height (Hm0) return periods. At five sites, the 1 in 100 year return period increased by over one meter. The enhanced dataset now includes more individual storm events and more complete records of storm events, providing a more robust foundation for coastal management and planning.

Development of an augmented reality (AR) glasses system for medical safety measures using artificial intelligence (AI)

Medical errors are not generally disclosed but they do happen and vary in severity. By prioritising patient safety, Dr Kohei Tanaka, a clinical engineer from the Tokyo University of Technology, wants to find a solution for preventing medical errors or, at the very least, ensuring they don’t escalate into serious incidents. The solution Tanaka is working on combines artificial intelligence (AI) with augmented reality (AR) glasses. The idea is that the glasses will provide medical professionals with accurate and reliable information at critical times. By providing them with rapid access to real-time data, the expectation is that this will reduce critical errors. As part of the validation process, newly qualified nurses will compare current training methods with training using his AR glasses and also complete a questionnaire on the viability of the glasses. So far, the researchers have been able to verify the retention of data provided through the glasses for a period of about three months and believe large-scale verification will reveal the mechanism of long-term memory in the future. By providing accurate, up-to-date information in a visual (still images and video) and auditory format, the glasses assist with the rapid delivery of critical data. However, there remain question marks regarding how much data can be effectively presented to the operator at the point of use as information overloading is to be avoided. The system included a voice-synthesised text reading guide as well as the display through the bone conduction earpiece within the device, which means the user can continue with the procedure while listening to the voice guide, reading reliance on visuals and thereby increasing safety.

Data retention in the light of copyright infringements: Protecting the status quo or seeking a third way? (Case C-470/21 La Quadrature du Net II)

In La Quadrature du Net II, the EU Court of Justice addressed – once again – the issue of the application of national retention legislation providing for the generalized retention of telecommunications data. This time, unlike in the previous cases, the problem was examined not in relation to measures used in the areas of combating serious crime or pursuing national security objectives, but in relation to counteracting online copyright infringements. Hence the case, decided by the CJEU sitting as the Full Court, concerned an important and topical issue, namely the possibility of countering mass copyright infringements in a situation where, according to a strict reading of the previous case law, the permissibility of imposing an obligation on telecommunications operators to collect information on users’ IP addresses was questioned. The LQDN II judgment – commented on by some as diverging from previous case law – is an important and necessary complement to the Court's earlier interpretation and, at the same time, a guideline on how to design retention laws to be applied not only in the area of countering copyright infringements, but also in other important areas, such as the fight against hate speech.

The evolution and formalization of anesthesia assistant roles across Canada.

The purpose of this Special Article is to document the evolution of the anesthesia assistant (AA) profession in Canada and summarize AA practice at Canadian institutions as it exists today, five decades after Quebec and 15years after most other provinces formalized AA practice. Through the Management Committee of the Association of Canadian University Departments of Anesthesia (ACUDA), we conducted a purposeful sampling of all ACUDA chairs or their delegates. We requested the following data: history of AAs becoming a reality in their particular province or region; potential recruitment pools; training programs and curricula; pathway to credentialing; funding, pay, retention, recruitment, and status of union representation; and metrics. Data were provided by 19 institutions in 8 provinces: Newfoundland and Labrador, Nova Scotia, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, and British Columbia. Given the different health care governance structures across the provinces, AA roles vary in terms of its associated technical, clinical, and educational responsibilities. The role of AAs in supporting anesthesia care through equipment maintenance and assistance with airway management, resuscitation, and administration of regional anesthesia seems to be well established, as is their role in providing brief intraoperative relief for anesthesiologists during a stable period of anesthesia. Anesthesia assistant duties continue to evolve, becoming more aligned with the specific institution and less dependent on the supervising anesthesiologist. Apart from the initial metrics collected during the Ontario ACT implementation pilot projects, we are not aware of any formal metrics, current or ongoing, being collected across Canada, related to either patient safety events or perioperative efficiency. This compilation of pan-Canadian AA data shows diverse models of practice and highlights the value to patients and the health care system as a whole of incorporating these allied professionals into the anesthesia care team (ACT). The present findings allow us to offer suggestions for consideration during discussions of retention, recruitment, program expansion, and cross-country collection of metrics and other data. We conclude by making six recommendations: 1. recognize that implementation of ACTs is a key element in solving the challenge of an increasing surgical backlog; 2. develop, or facilitate the development of, metrics and increase data-sharing nationally to enable health care authorities to better understand the importance of AAs in patient safety and perioperative efficiency; 3. develop and implement funding strategies to lower the barriers to AA training such as hospital-sponsored positions, ongoing salary support, and return-of-service arrangements; 4. ensure that salaries appropriately reflect the increased level of training and added levels of responsibility of certified AAs; 5. develop long-term strategies to ensure stable funding, recruitment and retention, and a better match between the number of AA training positions and the need for newly certified AAs; and 6. engage all stakeholders to acknowledge that AAs, as knowledgeable and specifically trained assistants, not only fulfill their defined clinical role but also contribute significantly to patient safety and clinical efficiency by assuming nondirect patient care tasks.

Radiation Anomaly Detection of Sub-Band Optical Remote Sensing Images Based on Multiscale Deep Dynamic Fusion and Adaptive Optimization

Radiation anomalies in optical remote sensing images frequently occur due to electronic issues within the image sensor or data transmission errors. These radiation anomalies can be categorized into several types, including CCD, StripeNoise, RandomCode1, RandomCode2, ImageMissing, and Tap. To ensure the retention of image data with minimal radiation issues as much as possible, this paper adopts a self-made radiation dataset and proposes a FlexVisionNet-YOLO network to detect radiation anomalies more accurately. Firstly, RepViT is used as the backbone network with a vision transformer architecture to better capture global and local features. Its multiscale feature fusion mechanism efficiently handles targets of different sizes and shapes, enhancing the detection ability for radiation anomalies. Secondly, a feature depth fusion network is proposed in the Feature Fusion part, which significantly improves the flexibility and accuracy of feature fusion and thus enhances the detection and classification performance of complex remote sensing images. Finally, Inner-CIoU is used in the Head part for edge regression, which significantly improves the localization accuracy by finely adjusting the target edges; Slide-Loss is used for classification loss, which enhances the classification robustness by dynamically adjusting the category probabilities and markedly improves the classification accuracy, especially in the sample imbalance dataset. Experimental results show that, compared to YOLOv8, the proposed FlexVisionNet-YOLO method improves precision, recall, mAP0.5, and mAP0.5:0.9 by 3.5%, 7.1%, 4.4%, and 13.6%, respectively. Its effectiveness in detecting radiation anomalies surpasses that of other models.

More is better? The role of strategic data management in a lean manufacturing process

Purpose Lean manufacturing has been pivotal in emphasizing the reduction of cycle times, minimizing manufacturing costs and diminishing inventories. This research endeavors to formulate a lean data management paradigm, through the design and execution of a strategic edge-cloud data governance approach. This study aims to discern anomalous or unforeseen patterns within data sets, enabling an efficacious examination of product shortcomings within manufacturing processes, while concurrently minimizing the redundancy associated with the storage, access and processing of nonvalue-adding data. Design/methodology/approach Adopting a lean data management framework within both edge and cloud computing contexts, this study ensures the preservation of significant time series sequences, while ascertaining the optimal quantity of normal time series data to retain. The efficacy of detected anomalous patterns, both at the edge and in the cloud, is assessed. A comparative analysis between traditional data management practices and the strategic edge-cloud data governance approach facilitates an exploration into the equilibrium between anomaly detection and space conservation in cloud environments for aggregated data analysis. Findings Evaluation of the proposed framework through a real-world inspection case study revealed its capability to navigate alternative strategies for harmonizing anomaly detection with data storage efficiency in cloud-based analysis. Contrary to the conventional belief that retaining comprehensive data in the cloud maximizes anomaly detection rates, our findings suggest that a strategic edge-cloud data governance model, which retains a specific subset of normal data, can achieve comparable or superior accuracy with less normal data relative to traditional methods. This approach further demonstrates enhanced space efficiency and mitigates various forms of waste, including temporal delays, storage of noncontributory normal data, costs associated with the analysis of such data and excess data transmission. Originality/value By treating inspected normal data as nonvalue-added, this study probes the intricacies of maintaining an optimal balance of such data in light of anomaly detection performance from aggregated data sets. Our proposed methodology augments existing research by integrating a strategic edge-cloud data governance model within a lean data analytical framework, thereby ensuring the retention of adequate data for effective anomaly detection.

Randomised pilot and feasibility trial of a group intervention for men who perpetrate intimate partner violence against women

BackgroundThere is a need for robust evidence on the effectiveness and cost-effectiveness of domestic abuse perpetrator programmes in reducing abusive behaviour and improving wellbeing for victim/survivors. While any randomised controlled trial can present difficulties in terms of recruitment and retention, conducting such a trial with domestic abuse perpetrators is particularly challenging. This paper reports the pilot and feasibility trial of a voluntary domestic abuse perpetrator group programme in the United Kingdom.MethodsThis was a pragmatic individually randomised pilot and feasibility trial with an integrated qualitative study in one site (covering three local-authority areas) in England. Male perpetrators were randomised to either the intervention or usual care. The intervention was a 23-week group programme for male perpetrators in heterosexual relationships, with an average of three one-to-one sessions, and one-to-one support for female current- or ex-partners delivered by third sector organisations. There was no active control treatment for men, and partners of control men were signposted towards domestic abuse support services. Data were collected at three-monthly intervals for nine months from male and female participants. The main objectives assessed were recruitment, randomisation, retention, data completeness, fidelity to the intervention model, and acceptability of the trial design.ResultsThis study recruited 36 men (22 randomly allocated to attend the intervention group programme, 14 to usual care), and 15 current- or ex-partners (39% of eligible partners). Retention and completeness of data were high: 67% of male (24/36), and 80% (12/15) of female participants completed the self-reported questionnaire at nine months. A framework for assessing fidelity to the intervention was developed. In interviews, men who completed all or most of the intervention gave positive feedback and reported changes in their own behaviour. Partners were also largely supportive of the trial and were positive about the intervention. Participants who were not allocated to the intervention group reported feeling disappointed but understood the rationale for the trial.ConclusionsIt was feasible to recruit, randomise and retain male perpetrators and female victim/survivors of abuse and collect self-reported outcome data. Participants were engaged in the intervention and reported positive benefits. The trial design was seen as acceptable.Trial registrationISRCTN71797549, submitted 03/08/2017, retrospectively registered 27/05/2022.

The Art of Dying in a Digital World: How Death is Represented in Digital Art

With the rise of computing, human rituals and cultural practices have increasingly interwoven with digital technologies. In this milieu, the cultural facets of death and art are particularly noteworthy. The realm of art has undergone a seismic shift due to computing, developing new artistic forms and enhancing the reach of older forms. Concurrently, aspects of death are now integrated into digital technologies, manifesting in unique ways such as the retention of data from the deceased, digital expressions of grief, online memorials, and the like. In light of this, our fundamental, qualitative, and exploratory study endeavors to map out and discuss digital art pieces that delve into themes of death, mourning, and digital afterlife. In doing so, we probe the intersection of art, death, and technology. Our findings delineate the primary methodologies and technologies employed in digital art centered around death, and the nuanced modalities through which this subject is portrayed in digital media.

Consent as a compositional act – a framework that provides clarity for the retention and use of data

BackgroundInformed consent is one of the key principles of conducting research involving humans. When research participants give consent, they perform an act in which they utter, write or otherwise provide an authorisation to somebody to do something. This paper proposes a new understanding of the informed consent as a compositional act. This conceptualisation departs from a modular conceptualisation of informed consent procedures.MethodsThis paper is a conceptual analysis that explores what consent is and what it does or does not do. It presents a framework that explores the basic elements of consent and breaks it down into its component parts. It analyses the consent act by first identifying its basic elements, namely: a) data subjects or legal representative that provides the authorisation of consent; b) a specific thing that is being consented to; and c) specific agent(s) to whom the consent is given.ResultsThis paper presents a framework that explores the basic elements of consent and breaks it down into its component parts. It goes beyond only providing choices to potential research participants; it explains the rationale of those choices or consenting acts that are taking place when speaking or writing an authorisation to do something to somebody.ConclusionsWe argue that by clearly differentiating the goals, the procedures of implementation, and what is being done or undone when one consent, one can better face the challenges of contemporary data-intensive biomedical research, particularly regarding the retention and use of data. Conceptualising consent as a compositional act enhances more efficient communication and accountability and, therefore, could enable more trustworthy acts of consent in biomedical science.

An Efficient Limited Memory Multi-Step Quasi-Newton Method

This paper is dedicated to the development of a novel class of quasi-Newton techniques tailored to address computational challenges posed by memory constraints. Such methodologies are commonly referred to as “limited” memory methods. The method proposed herein showcases adaptability by introducing a customizable memory parameter governing the retention of historical data in constructing the Hessian estimate matrix at each iterative stage. The search directions generated through this novel approach are derived from a modified version closely resembling the full memory multi-step BFGS update, incorporating limited memory computation for a singular term to approximate matrix–vector multiplication. Results from numerical experiments, exploring various parameter configurations, substantiate the enhanced efficiency of the proposed algorithm within the realm of limited memory quasi-Newton methodologies category.

Public information disclosure practices and civic participation: The roles of social justice perceptions and internet use

This study examined the relationship between public information disclosure practices and civic participation and the roles of social justice perceptions and internet use in that relationship. Our data comes from the 2019 Chinese Social Survey (CSS, 2019), a large nationwide household survey across 31 provinces (autonomous regions and municipalities). To ensure data relevance to our main variables, we conducted screening, resulting in the retention of data from 9 621 participants (female = 56 73%; mean age = 46 62 years, SD = 14 22 years). The results indicated that public information disclosure practices were associated with higher levels of civic participation. Specifically, public information disclosure practices were positively associated with institutional civic participation and negatively associated with non-institutional civic participation. Moreover, citizens' perception of social justice indirectly mediated civic participation. Internet use moderated the relationship between public information disclosure practices and the perception of social justice. These findings underscore the importance of types of public information disclosures in routine civil society governance. By implication, public information disclosures are tools for empowered civic participation with social justice

Multi-scale spatial features and temporal attention mechanisms: advancing the accuracy of ENSO prediction

The exponential progression in oceanic observational technology has fostered the accumulation of substantial time series data pivotal for predictions in ocean meteorology. Foremost among the phenomena observed is El Niño-Southern Oscillation (ENSO), a critical determinant in the interplay of global ocean atmosphere interactions, with its severe manifestations inducing extreme meteorological conditions. Therefore, precisely predicting ENSO events carries immense gravitas. Historically, predictions hinged primarily on dynamic models and statistical approaches; however, the intricate and multifaceted spatiotemporal dynamics of ENSO events have often impeded the accuracy of these traditional methodologies. A notable lacuna in contemporary research is the insufficient exploration of long-term dependencies within oceanic data and the suboptimal integration of spatial information derived from spatiotemporal data. To address these limitations, this study introduces a forward-thinking ENSO prediction framework synergizing multiscale spatial features with temporal attention mechanisms. This innovation facilitates a more profound exploration of temporal and spatial domains, enhancing the retention of extensive-period data while optimizing the use of spatial information. Preliminary analyses executed on the global ocean data assimilation system dataset attest to the superior efficacy of the proposed method, underscoring a substantial improvement over established methods including SA-convolutional long short-term memory, particularly in facilitating long-term predictions.The source code and datasets are provided. The code is available at https://github.com/tse1998/ENSO-prediction.

Biomechanical phenotyping pipeline for stalk lodging resistance in maize

Stalk lodging (structural failure crops prior to harvest) significantly reduces annual yields of vital grain crops. The lack of standardized, high throughput phenotyping methods capable of quantifying biomechanical plant traits prevents comprehensive understanding of the genetic architecture of stalk lodging resistance. A phenotyping pipeline developed to enable higher throughput biomechanical measurements of plant traits related to stalk lodging is presented. The methods were developed using principles from the fields of engineering mechanics and metrology and they enable retention of plant-specific data instead of averaging data across plots as is typical in most phenotyping studies. This pipeline was specifically designed to be implemented in large experimental studies and has been used to phenotype over 40,000 maize stalks. The pipeline includes both lab- and field-based phenotyping methodologies and enables the collection of metadata. Best practices learned by implementing this pipeline over the past three years are presented. The specific instruments (including model numbers and manufacturers) that work well for these methods are presented, however comparable instruments may be used in conjunction with these methods as seen fit.•Efficient methods to measure biomechanical traits and record metadata related to stalk lodging.•Can be used in studies with large sample sizes (i.e., > 1,000).

Model of collaborative organizational learning using soft systems and storytelling

Collaborative learning in the organization promotes interaction among participants to achieve common learning objectives and shared goals, based on peer interdependence, personal and group responsibility, skills and teamwork, through direct communication and purposeful reflection of members. In this context, this paper proposes a model of collaborative learning using the narrative provided by the storytelling as a mechanism for externalizing tacit knowledge in complex organizations. Storytelling is considered as an important tool, since stories have the virtue of connecting participants emotionally, providing them with a shared meaning in the concepts to be analysed, and stimulating reflection, creativity, memorization and retention of data in the resolution of complex problems. Relevant factors of collaborative learning and storytelling were developed and also the role of the actors and readers participating in the story and the context. A methodology of soft systems was implemented, as a problem-solving approach that aims to understand and address the complexity of human systems, in this case collaborative learning groups in the organization. Factors influencing the learning process were identified. The soft systems approach helps participants understand the multifaceted nature of educational challenges and consider a variety of perspectives and factors that influence the learning process. A collaborative learning method using storytelling and soft system methodology was used that brings the organizational professional closer to the reality of strategic business management, sectioning three case studies: a company from the productive sector, another from the services sector and a start-up company.

Supra-Quantum Dot Assemblies to Maximize Color-Based Multiplexed Fluorescence Detection with a Smartphone Camera.

Photoluminescence (PL) imaging and bioanalysis with smartphone-based devices are of growing interest for point-of-care/point-of-need diagnostics. Strategies for maximizing sensitivity have been explored in this context, but color multiplexing has been very limited, with its maximum level unexplored. Here, we evaluated color multiplexing with smartphone-based PL imaging by using supra-nanoparticle assemblies of quantum dots (supra-QDs). These materials were prepared as composite colors that were tailored to the red-green-blue (RGB) color space of smartphone cameras by coassembling different ratios of R-, G-, and B-emitting QDs on a silica nanoparticle scaffold. The supra-QDs were characterized and used to label cell-sized objects that were measured under flow with a smartphone-based device. Each color followed an approximately linear trajectory in the RGB space, and training of support vector machine models enabled color classification with overall accuracies ≥87% for 10-color multiplexing and better accuracies for fewer colors. Most misclassification occurred at low signal levels, such that establishing a nonclassifiable zone near the origin of RGB color space improved the overall 10-color classification accuracy to ≥94%. Similar improvements in accuracy with greater retention of data were possible with a probabilistic rather than a radial threshold. Simulations that were parameterized by experimental data suggested that ≥14-color multiplexing with accuracies ≥90% should be possible with an optimized supra-QD color set. This study is an important foundation for advancing RGB color-based multiplexing for imaging and analyses with smartphone cameras and related charge-coupled device and CMOS color image sensor technologies.

An Identity Privacy-Preserving Scheme against Insider Logistics Data Leakage Based on One-Time-Use Accounts

Digital transformation of the logistics industry triggered by the widespread use of Internet of Things (IoT) technology has prompted a significant revolution in logistics companies, further bringing huge dividends to society. However, the concurrent accelerated growth of logistics companies also significantly hinders the safeguarding of individual privacy. Digital identity has ascended to having the status of a prevalent privacy-protection solution, principally due to its efficacy in mitigating privacy compromises. However, the extant schemes fall short of addressing the issue of privacy breaches engendered by insider maleficence. This paper proposes an innovative identity privacy-preserving scheme aimed at addressing the quandary of internal data breaches. In this scheme, the identity provider furnishes one-time-use accounts for logistics users, thereby obviating the protracted retention of logistics data within the internal database. The scheme also employs ciphertext policy attribute-based encryption (CP-ABE) to encrypt address nodes, wherein the access privileges accorded to logistics companies are circumscribed. Therefore, internal logistics staff have to secure unequivocal authorization from users prior to accessing identity-specific data and privacy protection of user information is also concomitantly strengthened. Crucially, this scheme ameliorates internal privacy concerns, rendering it infeasible for internal interlopers to correlate the users’ authentic identities with their digital wallets. Finally, the effectiveness and reliability of the scheme are demonstrated through simulation experiments and discussions of security.

Obligations of the providers of electronic communications services in terms of the general and indiscriminate retention of traffic and location data – a commentary on the Judgment of the Court (Grand Chamber) of 5 April 2022 in the Case C-140/20 G.D. v The Commissioner of the Garda Síochána and Others.

Over the years, the issue of indiscriminate and general retention of traffic and location data has been considered many times by the Court of Justice. The Court has concluded that this is a tool that violates the rights included in the Charter of Fundamental Rights and that its further use should be limited only to the most serious purposes. In the commented judgment, the right of Member States to impose obligation of indiscriminate and general retention of data of users of this network onto the providers of electronic communications services was limited to precisely defined situations and in certain situations, methods of data retention that are less infringing on fundamental rights were explicitly named and allowed by the Court.

MAGNETO and DeepInsight: Extended Image Translation with Semantic Relationships for Classifying Attack Data with Machine Learning Models

The translation of traffic flow data into images for the purposes of classification in machine learning tasks has been extensively explored in recent years. However, the method of translation has a significant impact on the success of such attempts. In 2019, a method called DeepInsight was developed to translate genetic information into images. It was then adopted in 2021 for the purpose of translating network traffic into images, allowing the retention of semantic data about the relationships between features, in a model called MAGNETO. In this paper, we explore and extend this research, using the MAGNETO algorithm on three new intrusion detection datasets—CICDDoS2019, 5G-NIDD, and BOT-IoT—and also extend this method into the realm of multiclass classification tasks using first a One versus Rest model, followed by a full multiclass classification task, using multiple new classifiers for comparison against the CNNs implemented by the original MAGNETO model. We have also undertaken comparative experiments on the original MAGNETO datasets, CICIDS17, KDD99, and UNSW-NB15, as well as a comparison for other state-of-the-art models using the NSL-KDD dataset. The results show that the MAGNETO algorithm and the DeepInsight translation method, without the use of data augmentation, offer a significant boost to accuracy when classifying network traffic data. Our research also shows the effectiveness of Decision Tree and Random Forest classifiers on this type of data. Further research into the potential for real-time execution is needed to explore the possibilities for extending this method of translation into real-world scenarios.

Privacy-preserving continual learning methods for medical image classification: a comparative analysis.

The implementation of deep learning models for medical image classification poses significant challenges, including gradual performance degradation and limited adaptability to new diseases. However, frequent retraining of models is unfeasible and raises concerns about healthcare privacy due to the retention of prior patient data. To address these issues, this study investigated privacy-preserving continual learning methods as an alternative solution. We evaluated twelve privacy-preserving non-storage continual learning algorithms based deep learning models for classifying retinal diseases from public optical coherence tomography (OCT) images, in a class-incremental learning scenario. The OCT dataset comprises 108,309 OCT images. Its classes include normal (47.21%), drusen (7.96%), choroidal neovascularization (CNV) (34.35%), and diabetic macular edema (DME) (10.48%). Each class consisted of 250 testing images. For continuous training, the first task involved CNV and normal classes, the second task focused on DME class, and the third task included drusen class. All selected algorithms were further experimented with different training sequence combinations. The final model's average class accuracy was measured. The performance of the joint model obtained through retraining and the original finetune model without continual learning algorithms were compared. Additionally, a publicly available medical dataset for colon cancer detection based on histology slides was selected as a proof of concept, while the CIFAR10 dataset was included as the continual learning benchmark. Among the continual learning algorithms, Brain-inspired-replay (BIR) outperformed the others in the continual learning-based classification of retinal diseases from OCT images, achieving an accuracy of 62.00% (95% confidence interval: 59.36-64.64%), with consistent top performance observed in different training sequences. For colon cancer histology classification, Efficient Feature Transformations (EFT) attained the highest accuracy of 66.82% (95% confidence interval: 64.23-69.42%). In comparison, the joint model achieved accuracies of 90.76% and 89.28%, respectively. The finetune model demonstrated catastrophic forgetting in both datasets. Although the joint retraining model exhibited superior performance, continual learning holds promise in mitigating catastrophic forgetting and facilitating continual model updates while preserving privacy in healthcare deep learning models. Thus, it presents a highly promising solution for the long-term clinical deployment of such models.

Prioritizing Personal Data Protection in Insurance Organizations

This literature review focuses on the importance of prioritizing personal data security in insurance organizations in the context of Web 2.0 and the fourth industrial revolution. With the increasing use of digital transactions and data sharing, there is a need for improved data security postures to mitigate vulnerability issues. Personal data is a valuable asset that is targeted by malicious individuals, which highlights the importance of establishing proactive and trustworthy security frameworks. The study evaluates the security threats associated with the collection, processing, storage, and retention of personal data in insurance organizations, and examines the legal and technical provisions for its protection. Through an analytical review methodology, this paper highlights the urgent need to establish robust data security measures to ensure that personal data is secure when held in the custody of insurance organizations and personal management information systems.