General Data Research Articles

Analysis of Grip Strength and Its Explanatory Factors in Older Patients with Gastrointestinal Tumours.

To investigate the grip strength of older patients with gastrointestinal tumours and analyse its explanatory factors. A total of 170 older patients with gastrointestinal tumours admitted to the Gastrointestinal Surgery Department of a Grade-III hospital in the Zhejiang province from January 2022 to December 2022 were selected as the investigation participants. Among them, there were 102 cases of gastric cancer (60.0%) and 68 cases of colon cancer (40.0%). There were 110 male patients (64.7%) and 60 female patients (35.3%), with patient ages ranging between 64 and 82 years old. A cross-sectional survey was conducted using a general data questionnaire, grip apparatus and the Nutritional Risk Screening 2002 (NRS-2002) score. Multiple linear regression was used to analyse the explanatory factors on the grip strength of older patients with gastrointestinal tumours. The grip strength of older patients with gastrointestinal tumours was 25.3±5.3 kg. Multiple linear regression analysis showed that gender, age, disease, body mass index (BMI), albumin concentration and the NRS-2002 score were the main explanatory factors on grip strength (P<0.05). Grip strength was lower in older patients with gastrointestinal tumours and was lower in females, those aged ≥70 years, patients with colorectal cancer, individuals with a BMI of <18.5, those with an albumin concentration of <35g/l and those with an NRS2002 score of ≥3. Clinical staff should dynamically evaluate the level of grip strength in patients with gastrointestinal tumors and develop individualized interventions based on the related explanatory factors.

Effect of prenatal glucocorticoid therapy on hearing screening in premature infants

Objective:To explore the effect of prenatal glucocorticoids therapy on hearing screening in premature infants Methods:Data of 693 preterm infants with gestational age of 24-34+6weeks admitted to theJiangxi Maternal and Child Health Hospital within 24 h after birth from June 2022 to June 2023 were retrospectively analyzed. The infants were divided into the DXM group （544 cases） and the non-DXM group （149 cases） based on whether dexamethasone （DXM） was administered prenatally. General data of preterm infants and parturients in two groups were compared, and the effects of different doses and timing of DXM on hearing screening were analyzed. Results:In the terms of preliminary hearing screening. the pass rate of initial hearing screening in DXM group was significantly higher than that in non-DXM group（53.9% vs 35.6%）, with statistical significance（P<0.05）. Further subgroup analysis showed that the passing rate of preliminary hearing screening in adequate prenatal dose（=4 doses） DXM group（58.1%） was significantly higher than that in insufficient group（48.0%） and excessive group（42.4%）, with statistical significance（P<0.05）. Administering DXM 48 hours to 7 days before birth resulted in a higher pass rate for initial hearing screening compared to administration <48 hours or >7 days before birth （56.4% vs. 48.6%）, with a statistically significant difference （P < 0.05）. In terms of re-hearing screening, the pass rate of secondary hearing screening was not significantly correlated with DXM treatment（P>0.05）, but was significantly correlated with gestational age, birth weight, hospital stays, invasive mechanical ventilation, and common neonatal diseases（bronchopulmonary dysplasia, respiratory distress syndrome）（P<0.05）. Among them, bronchopulmonary dysplasia was an independent risk factor forsecondary hearing screening referral（P<0.05）. Conclusion:A single course of adequate dexamethasone use within 48 h-7 d of prenatal has a positive effect on the preliminary hearing screening of preterm infants.

Applications of Medical Big Data and Personal Information Protection

In the current social environment, the generation and application of medical big data is closely related to the personal information security of every citizen. With the continuous development of medical technology, the application of medical big data has become increasingly widespread. On one hand, it can promote the development of medical technology, including drug development, personalized medicine, and precision treatment. On the other hand, medical big data also concerns the private information of citizens, which is a crucial part of personal information protection. However, in practice, there often exists a contradiction between medical big data and the protection of citizens' personal information rights. From the perspective of medical big data, there are overlaps and differences among citizens' personal information, privacy information, medical health information, and medical big data information. Relying solely on the “informed consent” rule for personal information cannot effectively protect personal information comprehensively or guarantee the application and development of medical big data. This article analyzes the issues surrounding Chinese medical big data in terms of personal information from a legal perspective, and proposes corresponding solutions to these issues. It is hoped that this research can contribute to advancing the legalization of medical data openness, re-examining patients' personal information in the context of medical big data, and balancing the conflicts between individual privacy interests, public interests, and economic benefits of medical data, thus better protecting individuals' personal information interests under social and private benefits.

ARTIFICIAL INTELLIGENCE: LEGAL IMPLICATIONS AND CHALLENGES

Artificial intelligence (AI) technologies are rapidly transforming various aspects of society, from healthcare and finance to transportation and education. While AI offers tremendous potential for innovation and efficiency, its widespread adoption raises significant legal implications and challenges. This paper examines the legal landscape surrounding AI, focusing on key areas such as privacy, liability, intellectual property, and employment law. One of the primary concerns with AI is the privacy implications stemming from the collection, storage, and analysis of vast amounts of data. Regulations such as the General Data Protection Regulation (GDPR) in Europe and the California Consumer Privacy Act (CCPA) in the United States aim to safeguard individuals' privacy rights and impose strict requirements on data handling practices. Another critical area of concern is liability, particularly regarding the accountability for AI-driven decisions that may result in harm to individuals or entities. Questions arise about who should be held responsible for such decisions the developers, users, or the AI systems themselves. Furthermore, AI raises complex issues related to intellectual property, including the ownership of AI-generated works, patentability of AI algorithms, and the protection of AI innovations. Additionally, the integration of AI into the workforce raises questions about the future of employment, job displacement, and the need for new regulations to protect workers in the age of automation. In conclusion, while AI presents unprecedented opportunities for advancement, it also poses significant legal challenges that require careful consideration and proactive regulation. Policymakers, legal professionals, and stakeholders must collaborate to develop frameworks that promote the responsible development, deployment, and regulation of AI technologies while safeguarding individual rights, privacy, and societal values.

Unveiling the Black Box: Bringing Algorithmic Transparency to AI

Overall, algorithmic transparency is an important aspect of responsible AI development and deployment. Ensuring that AI systems are transparent and accountable will help build trust and confidence in these systems and ensure that they are used ethically and effectively. Artificial intelligence (AI) has emerged as a cutting-edge domain that is fundamentally redefining different areas of daily experiences, such as health care, transport, finance, education, and others. The systems are not created for making a judgment like human judgment of natural language, spotting patterns and problem-solving; rather AI produces machines that also have intelligence level same as that of human beings. AI having more influence over us, it is to be considered the ethical directions of these tools and see that they operate under principles of transparency and accountability. The element regarding algorithmic transparency, which means the process of understanding the functioning and explanation of how AI systems make their decisions is the one that is most crucial. The issue of algorithm transparency is of fundamental importance for many considerations. AI systems are not only supported by fairness but also by their non-discrimination. If we do not know how a system of AI arrives at the decisions made, it becomes impossible to determine if the provided results meet equal treatment for everybody. If used in delicate areas like recruitment, credit, and legal system- where the AI-machine must make choices which are life changing, then this aspect is very important. On top of fairness, algorithmic transparency is also an important factor for accountability. If we are ignorant about what an artificial intelligence algorithm does and what is the source of its decision-making process, we are unable to track and classify the mistakes or mishaps of the system. This has always mattered when central to the operation of systems with high stake, such as those used in self-driving vehicles or in health care. Algorithmic transparency may be reached using different instruments. The transparent AI systems can be made by a more transparent design, for example, the simple modelling tools, that use interpretable models. Another method is designing technologies and techniques that can help people why the artificial systems difficult to be decoded but easy to understand which they can utilize in making decisions. Therefore, algorithmic transparency is a key factor of the AI made responsibly and used by the society. It is crucial that AI machines are both transparent and accountable since this will lead to people building trust in the system and accepting its ethical and practical implications. This paper examines regulation of algorithmic transparency in the EU, specifically provisions under the General Data Protection Regulation (GDPR), it aims to situate analysis of the GDPR's provisions on explainability of AI systems within broader technology ethics and policy discourse. The paper's scope is limited to EU regulations applicable to AI data processing transparency.

Economic Consequences of the Right to be Forgotten

The right to privacy and control over one’s own information entails the legal right to request the removal of search results relating to one’s personality if all of those results do not serve the originality of the information or the processing purposes. This paper’s objective is to examine the case for a right to be forgotten from an economic standpoint and the implications of satisfying millions of requests, as well as how the General Data Protection Regulation (GDPR) has changed the way that this right is justified in society. This paper examines the economic consequences of reformulating requirements for realization and the right to remove previously classified material. The consequences are directly related to: the losses caused by a reduction in the amount of data accessible via search engines and the expected reduction in the operators’ commercial interest; the way the right to be forgotten is reflected in the costs associated with processing requests; and the negative effects of violations and sanctions. The conclusions will also evaluate the effects of each search engine’s obligations under the new territorial extension, as well as whether the benefits of upholding this human right outweigh all costs and efforts.

It is not just a click: protection of in-game behaviour data as biometric data under the GDPR

This article investigates the potential classification of players’ in-game behaviour data from basic computer peripherals, like keyboards and mouses (behavioural telemetry), as biometric data under the General Data Protection Regulation (GDPR). Modern technologies allow these types of data to reveal a wealth of personal information, including age, gender, emotional states, collaboration abilities, problem-solving skills, economic status, consumption habits, and character traits. The analysis is structured into three sections: understanding the nature of in-game behaviour data, assessing its alignment with the GDPR’s definition of biometric data, and examining the repercussions for gaming companies if such data is deemed biometric. The conclusion asserts that behavioural telemetry and gameplay metrics can indeed be considered biometric under the GDPR, as it plays a crucial role in player identification and exhibits similarities with biometric systems in the context of game analytics processes. This classification implies heightened legal responsibilities for gaming companies, triggering risk-based provisions within the GDPR. The article introduces a novel perspective by suggesting that seemingly innocuous in-game behaviour data may fall under the purview of biometric data.

Endoscopic ultrasound-guided biliary drainage vs percutaneous transhepatic bile duct drainage in the management of malignant obstructive jaundice.

Malignant obstructive jaundice (MOJ) is a condition characterized by varying degrees of bile duct stenosis and obstruction, accompanied by the progressive development of malignant tumors, leading to high morbidity and mortality rates. Currently, the two most commonly employed methods for its management are percutaneous transhepatic bile duct drainage (PTBD) and endoscopic ultrasound-guided biliary drainage (EUS-BD). While both methods have demonstrated favorable outcomes, additional research needs to be performed to determine their relative efficacy. To compare the therapeutic effectiveness of EUS-BD and PTBD in treating MOJ. This retrospective analysis, conducted between September 2015 and April 2023 at The Third Affiliated Hospital of Soochow University (The First People's Hospital of Changzhou), involved 68 patients with MOJ. The patients were divided into two groups on the basis of surgical procedure received: EUS-BD subgroup (n = 33) and PTBD subgroup (n = 35). Variables such as general data, preoperative and postoperative indices, blood routine, liver function indices, myocardial function indices, operative success rate, clinical effectiveness, and complication rate were analyzed and compared between the subgroups. In the EUS-BD subgroup, hospital stay duration, bile drainage volume, effective catheter time, and clinical effectiveness rate were superior to those in the PTBD subgroup, although the differences were not statistically significant (P > 0.05). The puncture time for the EUS-BD subgroup was shorter than that for the PTBD subgroup (P < 0.05). Postoperative blood routine, liver function index, and myocardial function index in the EUS-BD subgroup were significantly lower than those in the PTBD subgroup (P < 0.05). Additionally, the complication rate in the EUS-BD subgroup was lower than in the PTBD subgroup (P < 0.05). EUS-BD may reduce the number of punctures, improve liver and myocardial functions, alleviate traumatic stress, and decrease complication rates in MOJ treatment.

Identifying the risk factors for pancreatic fistula after laparoscopic pancreaticoduodenectomy in patients with pancreatic cancer.

Laparoscopic pancreaticoduodenectomy (LPD) is a surgical procedure for treating pancreatic cancer; however, the risk of complications remains high owing to the wide range of organs involved during the surgery and the difficulty of anastomosis. Pancreatic fistula (PF) is a major complication that not only increases the risk of postoperative infection and abdominal hemorrhage but may also cause multi-organ failure, which is a serious threat to the patient's life. This study hypothesized the risk factors for PF after LPD. To identify the risk factors for PF after laparoscopic pancreatoduodenectomy in patients with pancreatic cancer. We retrospectively analyzed the data of 201 patients admitted to the Fudan University Shanghai Cancer Center between August 2022 and August 2023 who underwent LPD for pancreatic cancer. On the basis of the PF's incidence (grades B and C), patients were categorized into the PF (n = 15) and non-PF groups (n = 186). Differences in general data, preoperative laboratory indicators, and surgery-related factors between the two groups were compared and analyzed using multifactorial logistic regression and receiver-operating characteristic (ROC) curve analyses. The proportions of males, combined hypertension, soft pancreatic texture, and pancreatic duct diameter ≤ 3 mm; surgery time; body mass index (BMI); and amylase (Am) level in the drainage fluid on the first postoperative day (Am > 1069 U/L) were greater in the PF group than in the non-PF group (P < 0.05), whereas the preoperative monocyte count in the PF group was lower than that in the non-PF group (all P < 0.05). The logistic regression analysis revealed that BMI > 24.91 kg/m² [odds ratio (OR) =13.978, 95% confidence interval (CI): 1.886-103.581], hypertension (OR = 8.484, 95%CI: 1.22-58.994), soft pancreatic texture (OR = 42.015, 95%CI: 5.698-309.782), and operation time > 414 min (OR = 15.41, 95%CI: 1.63-145.674) were risk factors for the development of PF after LPD for pancreatic cancer (all P < 0.05). The areas under the ROC curve for BMI, hypertension, soft pancreatic texture, and time prediction of PF surgery were 0.655, 0.661, 0.873, and 0.758, respectively. BMI (> 24.91 kg/m²), hypertension, soft pancreatic texture, and operation time (> 414 min) are considered to be the risk factors for postoperative PF.

Identifying the risk factors for pancreatic fistula after laparoscopic pancreaticoduodenectomy in patients with pancreatic cancer.

Laparoscopic pancreaticoduodenectomy (LPD) is a surgical procedure for treating pancreatic cancer; however, the risk of complications remains high owing to the wide range of organs involved during the surgery and the difficulty of anastomosis. Pancreatic fistula (PF) is a major complication that not only increases the risk of postoperative infection and abdominal hemorrhage but may also cause multi-organ failure, which is a serious threat to the patient's life. This study hypothesized the risk factors for PF after LPD. To identify the risk factors for PF after laparoscopic pancreatoduodenectomy in patients with pancreatic cancer. We retrospectively analyzed the data of 201 patients admitted to the Fudan University Shanghai Cancer Center between August 2022 and August 2023 who underwent LPD for pancreatic cancer. On the basis of the PF's incidence (grades B and C), patients were categorized into the PF (n = 15) and non-PF groups (n = 186). Differences in general data, preoperative laboratory indicators, and surgery-related factors between the two groups were compared and analyzed using multifactorial logistic regression and receiver-operating characteristic (ROC) curve analyses. The proportions of males, combined hypertension, soft pancreatic texture, and pancreatic duct diameter ≤ 3 mm; surgery time; body mass index (BMI); and amylase (Am) level in the drainage fluid on the first postoperative day (Am > 1069 U/L) were greater in the PF group than in the non-PF group (P < 0.05), whereas the preoperative monocyte count in the PF group was lower than that in the non-PF group (all P < 0.05). The logistic regression analysis revealed that BMI > 24.91 kg/m² [odds ratio (OR) =13.978, 95% confidence interval (CI): 1.886-103.581], hypertension (OR = 8.484, 95%CI: 1.22-58.994), soft pancreatic texture (OR = 42.015, 95%CI: 5.698-309.782), and operation time > 414 min (OR = 15.41, 95%CI: 1.63-145.674) were risk factors for the development of PF after LPD for pancreatic cancer (all P < 0.05). The areas under the ROC curve for BMI, hypertension, soft pancreatic texture, and time prediction of PF surgery were 0.655, 0.661, 0.873, and 0.758, respectively. BMI (> 24.91 kg/m²), hypertension, soft pancreatic texture, and operation time (> 414 min) are considered to be the risk factors for postoperative PF.

DATA WAREHOUSING IMPLEMENTATION IN HEDGE FUND ADMINISTRATION INDUSTRY

This research work tried to provide an outline specification for scope of implementation of data warehousing. This model is targeting industry as a whole and covering general system and data sets which are minimally required by each and every hedge fund administrator, however based on geographical location and changes in regulatory conditions, macroeconomic environment across the globe and software, networking and hardware technical advancements and security led to open doors for further study and research according changes in trends and scope as we discussed. Few areas identified can be classified as: Technical Scope, Financial and budgetary changes, Changes in regulatory process, Change in macroeconomic environments, Changes in accounting terms, Information Security. KEYWORDS: Data Warehousing, Hedge Funds, Extraction, Administration, Trading

A SAR Ship Detection Method Based on Adversarial Training.

SAR (synthetic aperture radar) ship detection is a hot topic due to the breadth of its application. However, limited by the volume of the SAR image, the generalization ability of the detector is low, which makes it difficult to adapt to new scenes. Although many data augmentation methods-for example, clipping, pasting, and mixing-are used, the accuracy is improved little. In order to solve this problem, the adversarial training is used for data generation in this paper. Perturbation is added to the SAR image to generate new samples for training, and it can make the detector learn more abundant features and promote the robustness of the detector. By separating batch normalization between clean samples and disturbed images, the performance degradation on clean samples is avoided. By simultaneously perturbing and selecting large losses of classification and location, it can keep the detector adaptable to more confrontational samples. The optimization efficiency and results are improved through K-step average perturbation and one-step gradient descent. The experiments on different detectors show that the proposed method achieves 8%, 10%, and 17% AP (Average Precision) improvement on the SSDD, SAR-Ship-Dataset, and AIR-SARShip, compared to the traditional data augmentation methods.

BRON: A blockchained framework for privacy information retrieval in human resource management

The correctness and the true validated data in Human Resource Management (HRM) are important for organizations as the data plays an impactful role in recruiting, developing, and retaining a skilled workforce. On one hand, the validated data in an organization helps in recruiting legitimate skillful employees; on the other hand, keeping the employee's data safe and maintaining privacy laws such as compliance with the General Data Protection Regulation (GDPR) is also an organization's responsibility. Besides, transparency in human resource management operations is crucial because it promotes trust and fairness within an organization. The present HRM systems are centralized in nature and their verifiable credential system is ineffective; this leads to the intentions of internal data sabotage or internal threats. Besides, the organizations' biases also become more prominent.In this paper, we address the above-mentioned problems with a blockchain framework for HRM to utilize the privacy of data access through a Privacy Information Retrieval (PIR) process. To be specific, our proposed framework called Blockchained piR of resOurces as humaN (BRON), is the first blockchain framework to show an effective mechanism to access data from organizations globally without hampering privacy. BRON uses a generalized user registration process to use the services of data access and in the background, it uses Zero-Knowledge Proofs (ZKPs) for global verification and PIR for privacy-based data retrieval. More specifically, credential verification and ZKP-based PIR are the highlights of our proposed BRON. Another interesting aspect of BRON is the use of Proof-of-Authority (PoA) to validate the anonymity and unlinkability of any HR operation. Finally, BRON has also contributed with a smart contract to incentivize the employees. BRON is very generic and easily be customizable as per the HR requirements. We run a set of experiments on BRON and observe that it is successful in providing privacy-assured data access and decentralized human resource data management. Overall, BRON provides 30% reduced latency and 35% better throughput as compared to the existing blockchain solutions in the direction of HRM.

Model of Shape Memory Alloy Actuator with the Usage of LSTM Neural Network.

Shape Memory Alloys (SMAs) are used to design actuators, which are one of the most fascinating applications of SMA. Usually, they are on-off actuators because, in the case of continuous actuators, the nonlinearity of their characteristics is the problem. The main problem, especially in control systems in these actuators, is a hysteretic loop. There are many models of hysteresis, but from a control theory point of view, they are not helpful. This study used an artificial neural network (ANN) to model the SMA actuator hysteresis. The ANN structure and training method are presented in the paper. Data were generated from the Preisach model for training. This approach allowed for quick and controllable data generation, making experiments thoroughly planned and repeatable. The advantage and disadvantage of this approach is the lack of disturbances. The paper's main goal is to model an SMA actuator. Additionally, it explores whether and how an ANN can describe and model the hysteresis loop. A literature review shows that ANNs are used to model hysteresis, but to a limited extent; this means that the hysteresis loop was modelled with a hysteretic element.

Effects of tramadol hydrochloride preemptive analgesia in kyphoplasty of thoracolumbar osteoporotic fractures under local anesthesia

To explore preemptive analgesic effect of preoperative intramural tramadol injection in percutaneous kyphoplasty (PKP) of vertebrae following local anesthesia. From August 2019 to June 2021, 118 patients with thoraco lumbar osteoporotic fractures were treated and divided into observation group and control group, with 59 patients in each gruop. In observation group, there were 26 males and 33 females, aged from 57 to 80 years old with an average of (67.69±4.75)years old;14 patients on T11, 12 patients on T12, 18 patients on L1, 15 patients on L2;tramadol with 100 mg was injected intramuscularly half an hour before surgery in observation group. In control group, there were 24 males and 35 females, aged from 55 to 77 years old with an average of (68.00±4.43) years old;19 patients on T11, 11 patients on T12, 17patients on L1, 12 patients on L2;the same amount of normal saline was injected intramuscularly in control group. Observation indicators included operation time, intraoperative bleeding, visual analogue scale (VAS) evaluation and recording of preoperative (T0), intraoperative puncture(T1), and working cannula placement (T2) between two groups of patients, at the time of balloon dilation (T3), when the bone cement was injected into the vertebral body (T4), 2 hours after the operation (T5), and the pain degree at the time of discharge(T6);adverse reactions such as dizziness, nausea and vomiting were observed and recorded;the record the patient's acceptance of repeat PKP surgery. All patients were successfully completed PKP via bilateral pedicle approach, and no intravenous sedative and analgesic drugs were used during the operation. There was no significant difference in preoperative general data and VAS(T0) between two groups (P>0.05). There was no significant difference in operation time and intraoperative blood loss between the two groups (P>0.05). VAS of T1, T2, T3, T4 and T5 in observation group were all lower than those in control group(P<0.05), and there was no significant difference in T6 VAS (P>0.05). T6 VAS between two groups were significantly lower than those of T0, and the difference was statistically significant (P<0.05). There was no significant difference in incidence of total adverse reactions between two groups (P>0.05). There was a statistically significant difference in the acceptance of repeat PKP surgery (P<0.05). Half an hour before operation, intramuscular injection of tramadol has a clear preemptive analgesic effect for PKP of single-segment thoracolumbar osteoporotic fracture vertebral body under local anesthesia, which could increase the comfort of patients during operation and 2 hours after operation, and improve patients satisfaction with surgery.

The association between vitamin D and the progression of diabetic nephropathy: insights into potential mechanisms.

Vitamin D deficiency (VDD) is prevalent in the population, with inadequate intake, impaired absorption and metabolism as the main causative factors. VDD increases the risk of developing chronic diseases such as type 2 diabetes mellitus (T2DM) and diabetic nephropathy (DN), but the molecular mechanisms underlying this phenomenon are not known. The aim of this study was to investigate the association and potential mechanisms of vitamin D levels with the progression of DN by analyzing general clinical data and using bioinformatics methods. The study included 567 diabetes mellitus type 2 (T2DM) patients from the Rocket Force Characteristic Medical Center as the case group and 221 healthy examinees as the normal control group. T2DM patients were categorized into T2DM, early diabetic nephropathy (EDN), and advanced diabetic nephropathy (ADN) based on the progression of diabetic nephropathy. The renal RNA-seq and scRNA-seq data of patients with DN were mined from public databases, and the differential expression of vitamin D-related genes in normal-EDN-ADN was analyzed by bioinformatics method, protein interaction network was constructed, immune infiltration was evaluated, single cell map was drawn, and potential mechanisms of VD and DN interaction were explored. Chi-square test showed that vitamin D level was significantly negatively correlated with DN progression (p < 0.001). Bioinformatics showed that the expression of vitamin D-related cytochrome P450 family genes was down-regulated, and TLR4 and other related inflammatory genes were abnormally up-regulated with the progression of DN. Vitamin D metabolism disturbance up-regulate "Nf-Kappa B signaling pathway," B cell receptor signaling pathway and other immune regulation and insulin resistance related pathways, and inhibit a variety of metabolic pathways. In addition, vitamin D metabolism disturbance are strongly associated with the development of diabetic cardiomyopathy and several neurological disease complications. VDD or vitamin D metabolism disturbance is positively associated with the severity of renal injury. The mechanisms may involve abnormal regulation of the immune system by vitamin D metabolism disturbance, metabolic suppression, upregulation of insulin resistance and inflammatory signalling pathways.

Synthetic data generation for a longitudinal cohort study – evaluation, method extension and reproduction of published data analysis results

Access to individual-level health data is essential for gaining new insights and advancing science. In particular, modern methods based on artificial intelligence rely on the availability of and access to large datasets. In the health sector, access to individual-level data is often challenging due to privacy concerns. A promising alternative is the generation of fully synthetic data, i.e., data generated through a randomised process that have similar statistical properties as the original data, but do not have a one-to-one correspondence with the original individual-level records. In this study, we use a state-of-the-art synthetic data generation method and perform in-depth quality analyses of the generated data for a specific use case in the field of nutrition. We demonstrate the need for careful analyses of synthetic data that go beyond descriptive statistics and provide valuable insights into how to realise the full potential of synthetic datasets. By extending the methods, but also by thoroughly analysing the effects of sampling from a trained model, we are able to largely reproduce significant real-world analysis results in the chosen use case.

Asynchronous Record Alignment of Network Flows for Incident Detection and Reconstruction

In today's interconnected digital landscape, the distribution of cyber threats presents a significant challenge to cyber security. Moreover, as of 2016, the amount of data in the world exceeds one zettabyte. Because of this, evidence-based network flow analytics is a critical component of modern network management and security. Problems such as anomalies in the network flow, cyber security incidents, alert generation, data pre-processing, network monitoring, network flow complexity, and data flow patterns become difficult to detect in massive network data flows. These specific problems can be addressed using Packet capture (PCAP). PCAP analysis is a standard network forensics process and investigation for assessing network behaviour and identifying anomalies. This work presents a method for analysing network flows for probable alignment of asynchronously recorded communications in heterogeneous networks. Using a proposed method for alignment, we can identify the relevant recordings aligned over two data streams for faster and more conclusive incident analysis. We use synthetic network incident scenarios for research experiments, detailing the generation of cyber event data and impact on cloud network traffic, followed by in-depth PCAP analysis. The automated cyber-attacks are simulated within a network infrastructure generating network flows in a PCAP format. Simulated cyber-attacks range from standard port scans, service scans, and specific scenarios like SQL injection, phishing, DoS or DDoS. We define analysis objectives and criteria for the in-depth PCAP analysis and alignment. The evidence gathered showcases valuable information about network data flow and its behaviour.

Exploring Trainees’ Behaviour in Hands-on Cybersecurity Exercises Through Data Mining

Despite the rising number of cybersecurity professionals, the demand for more experts in this field is still substantial. Cybersecurity professionals must also possess up-to-date knowledge and skills to counter cybersecurity threats’ dynamicity and rapidly evolving nature. Hands-on cybersecurity training is mandatory to practice various tools and improve one’s technical cybersecurity skills. Generally, an interactive learning environment is set, where trainees perform sophisticated tasks by accessing complete operating systems, applications, and networks. One of the main challenges that cybersecurity organizations are facing today is the generation of massive data through practice exercises. So, it becomes a problem to convert this data into knowledge to improve the overall quality of the learning system. The large amount of interaction data and its complexity also limit us to do automated analysis. Thus, these challenges for cybersecurity learners can be addressed through appropriate educational data analysis by having insights or testing hypotheses or models on a proper dataset. Revealing the patterns, rules, item sets and time taken by trainees while using any command line tool could help the trainer to assess the trainees and to provide feedback. Therefore, in this paper we are analyzing the frequency patterns and timing information captured from the trainees’ command line log to reveal their solving techniques, easy and struggling stages, slipups, and individual performance. Through our study, we aim to show how education and training providers can foresee learners who are less likely to succeed in a task or exhibit low performance, which can impede learning proficiency. With this knowledge, organizations and trainers can identify trainees who require additional attention or support. It may also be able to identify elements related to an organization like training aids, training methodology, etc. that need improvement. This study demonstrates the utility of data-mining techniques, specifically rule mining and sequential mining, to empower training designers to delve into datasets derived from cyber security training exercises.

Improving clinical abbreviation sense disambiguation using attention-based Bi-LSTM and hybrid balancing techniques in imbalanced datasets.

Clinical abbreviations pose a challenge for clinical decision support systems due to their ambiguity. Additionally, clinical datasets often suffer from class imbalance, hindering the classification of such data. This imbalance leads to classifiers with low accuracy and high error rates. Traditional feature-engineered models struggle with this task, and class imbalance is a known factor that reduces the performance of neural network techniques. This study proposes an attention-based bidirectional long short-term memory (Bi-LSTM) model to improve clinical abbreviation disambiguation in clinical documents. We aim to address the challenges of limited training data and class imbalance by employing data generation techniques like reverse substitution and data augmentation with synonym substitution. We utilise a Bi-LSTM classification model with an attention mechanism to disambiguate each abbreviation. The model's performance is evaluated based on accuracy for each abbreviation. To address the limitations of imbalanced data, we employ data generation techniques to create a more balanced dataset. The evaluation results demonstrate that our data balancing technique significantly improves the model's accuracy by 2.08%. Furthermore, the proposed attention-based Bi-LSTM model achieves an accuracy of 96.09% on the UMN dataset, outperforming state-of-the-art results. Deep neural network methods, particularly Bi-LSTM, offer promising alternatives to traditional feature-engineered models for clinical abbreviation disambiguation. By employing data generation techniques, we can address the challenges posed by limited-resource and imbalanced clinical datasets. This approach leads to a significant improvement in model accuracy for clinical abbreviation disambiguation tasks.