Перегляд за Автор "Kalashnikova, Larysa"

Зараз показуємо 1 - 2 з 2
  • Ескіз
    ДокументВідкритий доступ
    Artificial Intelligence in Bioengineering: Improving the Diagnosis and Treatment of Breast Cancer
    (КПІ ім. Ігоря Сікорського, 2024) Jiao Rui; Kalashnikova, Larysa
    48 sources were processed. Relevance: Medical artificial intelligence (AI) has evolved from assisted diagnosis to biotechnology, drug development, disease prediction, and health management, spanning four stages: auxiliary diagnosis, treatment, rehabilitation, and health management. It impacts smart hospital systems, chronic disease care, drug regulation, and epidemic monitoring, enhancing clinical diagnosis accuracy and flexibility while reducing reliance on human designers through autonomous learning. However, AI's "black box" nature and deep learning create risks, increasing medical uncertainties and legal disputes. The shift toward "AI decision-making with physician review" is reshaping medical practice. Article 49 of the Basic Medical Health Promotion Law (2020) supports AI's growth in healthcare, but its full benefits and risks remain uncertain, altering medical care duties as AI evolves. The purpose is to use artificial intelligence techniques to improve the efficiency and accuracy of medical diagnosis and treatment processes. The research object is to explore the potential applications and benefits of incorporating artificial intelligence technology into the field of bioengineering to enhance diagnostic and treatment processes. Tasks: 1. To review the existing literature on the application of Artificial Intelligence in bioengineering and healthcare. 2. To identify specific areas of bioengineering where Artificial Intelligence can be effectively utilised for diagnostic and therapeutic purposes. 3. To explore the capabilities of Artificial Intelligence techniques (e.g., machine learning and deep learning) in improving the accuracy and efficiency of diagnostic procedures. 4. To examine the potential impact of Artificial Intelligence on treatment planning, personalised medicine and patient prognosis in bioengineering. 5. To make recommendations for integrating Artificial Intelligence technologies into bioengineering practices to effectively improve diagnostic and therapeutic processes. The subjective of the research is the application of artificial intelligence in the field of enhanced medical diagnosis and treatment.
  • Ескіз
    ДокументВідкритий доступ
    Optimization of scanning parameters for CT and CBCT: a systematic review
    (КПІ ім. Ігоря Сікорського, 2025) Tsarenko, Mykola; Kalashnikova, Larysa
    Computed tomography (CT) and cone-beam computed tomography (CBCT) have revolutionized medical imaging by providing high-resolution, three-dimensional (3D) anatomical models for diagnostics, treatment planning, and surgical simulation. The accuracy of these models is highly dependent on scanning parameters such as slice thickness, spatial resolution, radiation dose, voltage, exposure time, and reconstruction algorithms. While optimized parameters can enhance image quality and segmentation accuracy, suboptimal settings may introduce artifacts, reduce anatomical fidelity, and compromise clinical outcomes [1]. CBCT is widely used in dentistry and maxillofacial surgery due to its lower radiation dose and high spatial resolution, whereas CT is preferred for comprehensive anatomical evaluations due to its superior soft tissue contrast [3]. The choice of scanning parameters requires balancing image clarity and patient safety. Studies have shown that an optimal slice thickness of 0.075–0.125 mm in CBCT and 0.5–1.25 mm in CT yields the best segmentation results [4]. Radiation dose must also be carefully adjusted; 0.1–0.3 mSv is typically sufficient for CBCT, while 2–5 mSv is recommended for CT [5]. Voltage settings of 80–100 kV (CBCT) and 100–120 kV (CT) help reduce beam hardening artifacts while maintaining contrast. Tube current should range between 4–10 mA for CBCT and 50–300 mA for CT to optimize noise reduction [6]. One of the major challenges in CT imaging is the presence of artifacts, including scatter artifacts, beam hardening artifacts, motion artifacts, and partial volume artifacts. Scatter artifacts degrade image quality due to unintended radiation deflection and can be mitigated using anti-scatter grids and beam collimation techniques [7]. Beam hardening artifacts, caused by differential X-ray absorption in dense structures, can be corrected using higher voltage settings and advanced reconstruction algorithms [4]. Motion artifacts, resulting from patient movement, can be minimized by reducing exposure time and employing motion correction software [3]. Partial volume artifacts, which affect the accuracy of tissue segmentation, can be addressed by reducing voxel size and applying high-pass filters. Traditional artifact reduction techniques such as high-pass filters, metal artifact reduction (MAR) algorithms, dual-energy CT (DECT), and Monte Carlo simulations have been widely implemented, but their effectiveness is often limited [8]. Recent advancements in Artificial Intelligence (AI)-based artifact correction have introduced new, data-driven methods that surpass conventional approaches in speed, accuracy, and adaptability [9]. This review provides a comprehensive analysis of CT and CBCT scanning parameters and typical artifacts, summarizing the optimal settings for different clinical applications. By refining scanning protocols and employing advanced artifact reduction techniques, the accuracy and reliability of anatomical models can be significantly improved, ensuring better diagnostic and therapeutic outcomes [10]