This scoping review examines how artificial intelligence (AI) can help reduce errors in radiology, an area where accuracy is critical to patient care. Radiology inherently involves complex image interpretation, and even minor mistakes can lead to delayed or wrong diagnoses, as well as inappropriate treatment. With advancements in AI, particularly in machine learning (ML), deep learning (DL), and natural language processing (NLP), there is growing interest in how these technologies can support radiologists and improve clinical outcomes. This review analyzed 12 studies that applied AI at various stages of the radiology workflow: before, during, and after image acquisition. AI has been used to assist in selecting appropriate imaging protocols, improving patient positioning, reducing motion artifacts, identifying abnormalities in scans, and supporting the generation of radiology reports. Across these applications, AI consistently demonstrated improvements in accuracy, sensitivity, and specificity, while significantly reducing reported error rates. In several cases, AI tools successfully flagged overlooked findings, acting as a safety net in high-pressure clinical environments. Despite these promising results, challenges remain. Issues such as algorithmic bias, limited data quality, and the need for robust clinical validation still require attention. Nonetheless, the evidence suggests that AI can serve as a valuable adjunct, enhancing diagnostic precision and supporting radiologists in delivering safer, more effective care. Overall, this review highlights the growing practical impact of AI in radiology and provides insights into how these technologies are already reshaping the field.
Keywords: artificial intelligence (ai); artificial intelligence in healthcare; artificial intelligence in radiology; error rates; scoping review.
Copyright © 2025, Velamala et al.