بهبودهای دقت تشخیصی برای سیستمهای تصویربرداری پزشکی با استفاده از هوش مصنوعی
کد: G-1824
نویسندگان: Mohammad Hossein Pourasad * ℗, Saleh Salehi Zahabi, Mohammad Dehghani, Somaye Norouzi
زمان بندی: زمان بندی نشده!
برچسب: پردازش سیگنال های پزشکی
دانلود: دانلود پوستر
خلاصه مقاله:
خلاصه مقاله
Introduction: The integration of Artificial Intelligence (AI) into medical image processing has significantly enhanced diagnostic accuracy. AI-powered Diagnostic Aid Systems (DAS) are transforming radiology, pathology, and other image-dependent specialties by automating detection, classification, and segmentation tasks. With increasing volumes of imaging data and a global shortage of expert radiologists, AI offers a scalable and efficient solution for improving diagnostic precision and supporting clinical decision-making. Methods: This study systematically reviewed peer-reviewed literature published between 2021 and 2025 on the application of AI in enhancing diagnosis accuracy in DAS for medical imaging. Various AI techniques, including convolutional neural networks (CNNs), deep learning, and transformer-based architectures, were evaluated. Comparative performance metrics (sensitivity, specificity, accuracy, and AUC) of AI models versus human experts and traditional rule-based systems were analyzed across modalities such as X-ray, MRI, CT, and digital pathology. Results: AI-enhanced DAS demonstrated superior accuracy compared to conventional systems across all studied modalities. For instance, deep learning models achieved a pooled sensitivity of 96.3% and specificity of 93.3% in digital pathology applications . In brain metastases detection using MRI, AI algorithms showed a detectability rate of 89% . Furthermore, AI tools in diagnostic imaging have been shown to reduce errors and accelerate diagnostic processes, leading to quicker patient diagnosis and reduced healthcare costs .arXiv Discussion: The results affirm AI's potential in enhancing diagnostic accuracy, particularly in settings with limited medical resources or high workloads. However, challenges such as data bias, model interpretability, and clinical integration persist. Standardizing model evaluation and promoting transparent AI systems are essential to gaining clinician trust and ensuring ethical deployment. Conclusion: AI-driven Diagnostic Aid Systems significantly improve diagnostic accuracy in medical image processing and hold promise for global healthcare enhancement. Future research should focus on real-world deployment, interoperability, and regulatory frameworks to fully harness AI’s potential in clinical practice.
کلمات کلیدی
AI;Diagnostic Aid Systems;Medical Medical Image;Deep Learning
