Fig 1 - uploaded by Walid Abushiba
Content may be subject to copyright.
Proposed framework for the integration of Explainable Artificial Intelligence with Large Language Models for threat detection.
Citations
... Patients' crucial data is being held hostage in a form of cyber blackmail, ultimately locking up entire healthcare systems and services. It is essential now more than ever to bolster these sectors with optimized cybersecurity strategies that can keep up with the ever growing sophistication of attacks and threats [2]. ...
The growing digitization of crucial infrastructure, especially in areas like health care, has put these sectors at risk of complex cyberattacks. Managing a patient's protected health information (PHI) and operational systems of health care facilities is highly sensitive; breaches can result in data theft, service interruptions, and even endanger the safety of patients. This paper presents the use of AI cybersecurity solutions for the protection of vital infrastructure and patient data with advanced anomaly detection, intrusion prevention, and threat intelligence systems. AI algorithms can analyze large amounts of information in real time and identify new threats while adapting defenses automatically. The study covers central applications of AI: machine learning-enhanced intrusion detection systems, endpoint protection, and behavior-based mitigation of cyber threats. It also tackles issues of model accuracy-false positives, privacy, and model size. Further, the paper presents an integrated cyberspace security approach designed specifically for the healthcare systems with the goals of minimizing exploitable assets and maximizing sensitive data security. The core directions are aimed at building transparent AI models and employing blockchain technology for improved assurance of data integrity.
... Deep learning applications, such as gas pipeline leakage detection, highlight AI's role in infrastructure surveillance, which can be extended to water, waste, and energy systems for sustainability [49]. Predictive analytics, already proven effective in healthcare optimization, can similarly improve environmental sector planning and management [50]. Fraud detection techniques in finance have been adapted to identify wasteful practices, ensuring that ESG resources are managed responsibly. ...
One of the key areas where Artificial Intelligence (AI) can counteract the forces of destruction and promote sustainability is intelligent decision-making, resource allocation, and minimizing environmental impact. This paper focuses on how AI aids environmental surveillance, energy management, waste management, and sustainable agriculture. AI-powered systems enhance climate modeling, energy grid optimization, supply chain efficiency, and recycling processes. By utilizing large datasets, real-time analytics, and AI, intelligent automation supports sustainability efforts globally. However, ethical concerns and energy consumption challenges should not deter AI from becoming a driving force in a greener future, especially with emerging technologies like blockchain and IoT. The objective of this paper is to explore AI's role in sustainable development through its applications, challenges, and future directions, demonstrating how AI-driven solutions can create a more resilient and environmentally friendly world.
... Sufian et al. (2024) emphasize that meticulous calculation of this angle is very critical to orthodontic practice. Other clinical studies further stress the necessity of having the right angle NLA during orthodontic therapy in order to advance aesthetic (Ghazal et al., 2024) and global patient acceptance. ...
Class II maloclussion is the most prevalent in orthodontic patients. During camolflauge treatment of this maloclussionnasolabial angel increases inevitably for which a threshold value needs to be defined. Objective: The objective of this study was to calculate the mean score for the modified profile of a woman of class II div 1, by digitally simulating a rise in nasolabial angle from the initial image. Methods: This cross-sectional study was undertaken at Puniab Dental Hospital/de' Montmorency College of Dentistry from July 15 to December 01, 2024. A profile picture and lateral cephalometric radiograph of a female with an untreated skeletal Class 2 Division I relationship, a normal mandibular plane angle and normal face height were used. The NLA of the subject's profile image was adjusted to 104.9±4º using Adobe Photoshop CS2. The base image was then digitally changed to produce additional profile photographs, imitating increase in nasolabial angle by 2.0, 4.0, and 6.0 standard deviations (corresponding images called C, B, and A in the questionnaire). Results: The mean age of lay persons was 29.14±5.41 years, minimum age was 18 and maximum age was 41 years. The gender of 90(58.1%) were males and 65(41.9%) were females. Mean attractiveness of facial profile was evaluated by calculating mean attractiveness as the lay people rank the images from 1 to 5. Mean attractiveness score was 4.74±0.44 for image B followed by 4.54±0.50 for image A, 4.37±0.50 for image C and 3.27±0.45 for base image. Conclusions: According to the study, both the nontreated and profile with biggest nasolabial angle (NLA) had the least pleasing appearance. To achieve an aesthetic profile at the end of treatment while treating a class II DIV 1 patient the nasolabial angle should not exceed 121°.
... Certain challenges persist especially regarding the auditability and transparency of AI systems for important healthcare operations [23]. Moreover, the aspects concerning data protection, consent, and the ethical boundaries of the use of AI in the medicine field are still extensively searched and debated [24]. ...
The use of Artificial Intelligence, especially machine learning and artificial neural networks, has dramatically increased in urology, assisting in innovative ways for diagnosis, prognosis, and treatment planning. This paper presents an up-to-date review of AI advances in the field of urology that pertain to its imaging aspects, particularly regarding the diagnosis of prostate cancer, kidney stones, and bladder cancer. The deep learning methods, especially convolutional neural networks, proved to be very effective in many medical imaging tasks, such as automated abnormal growth detection, organ segmentation, etc. Additionally, deep learning systems have performed well in predicting a patient’s outcome, including post-operative complications and recovery. Nevertheless, the progress made greatly differs from the goals set, and AI’s integration into clinical practice remains an unmet need due to obstacles posed by inefficient datasets and the opacity of some AI algorithms.This paper also discusses the key challenges in implementing AI tools in urology, as well as the potential for future research to enhance the accuracy, interpretability, and clinical applicability of AI-driven solutions. Ultimately, AI is poised to play a transformative role in urology, offering the potential for more personalized, efficient, and precise patient care.
The security of the Web is a significant issue for personal, corporate, and state users in the context of digitalisation. With all kinds of activities related to the Internet growing, different types of threats also emerge, including phishing, malware, ransomware, and others which threaten personal information, funds, and critical system structures. The present paper discusses the principal threats that are inherent in the web environment, the effects of these threats, and protection means. Phishing is a common kind of social engineering that aims at making the target release relevant information. Viruses and worms, in their broad sense, include Malware, which sneaks into systems to corrupt, steal or delete important information. Ransomware is a virus that encrypts a victim"s files and asks for payment for the decryption key. At the same time, drive-by downloads are another virus that installs themselves on a victim"s computer from compromised websites without the victim knowing. With spoofing and man-in-the-middle attacks, data integrity is not preserved. At the same time, SQL injection and cross-site scripting are aimed at controlling web applications in order to control databases. Ddos-attack or Distributed Denial of Service attacks on services knock them off balance by flooding them with traffic. Minimising risks entails keeping oneself posted on the latest developments, updating the software, locking passwords in the cyber world, using a two-factor identification key, and making sure that different strains of technology have backup copies of the materials that have been tampered with. Adhering to security practices and being alert to new threats is vital for an organisation to have a safe online existence. Hence, this study emphasises the need to adopt appropriate measures for evaluating web threats in order to have safe interactions.
