Wedoany.com Report on Feb 6th, The integration of artificial intelligence with structured reporting has brought significant efficiency improvements to radiology workflows. A recent study shows that this combination not only enhances diagnostic accuracy but also substantially shortens report turnaround times. The related analysis has been published in RSNA's flagship journal, Radiology.
Dr. Robert M. Siepmann and his team from the Department of Diagnostic and Interventional Radiology at University Hospital Aachen in Switzerland stated: "With the growing demand for radiology services and increasing complexity of imaging, it is crucial to develop reporting workflows that balance efficiency and accuracy while reducing non-image interactions. However, the impact of different reporting modes and AI assistance on the diagnostic process remains unclear."
The research team conducted a quantitative analysis to explore the effects of different reporting modes on radiologists' focus, accuracy, and efficiency. Eight participants (including four novices and four more experienced physicians) evaluated 35 bedside chest X-rays using three reporting modes: free-text reports, structured reports, and AI-pre-filled structured reports. The study employed an eye-tracking system to monitor the frequency of gaze shifts.
The results showed that free-text and structured reports had similar diagnostic accuracy, but accuracy significantly improved when combined with AI pre-filling. The average reporting time decreased from 88 seconds for free-text reports to 37 seconds for structured reports, and further shortened to 25 seconds for the AI combination. When using AI, rapid eye movements decreased, and total fixation duration was reduced.
All participants preferred the combination of AI and structured reporting, with seven specifically highlighting its efficiency advantages. The research team concluded: "Structured reporting enhances efficiency by guiding visual attention, which is particularly beneficial for less experienced physicians. AI-pre-filled structured reports further improve diagnostic accuracy. Future research should focus on factors such as the timing of AI output, framework, and interface design to optimize human-computer collaboration."
This study confirms that the combination of structured reporting and AI holds significant value in radiology applications, effectively enhancing workflow efficiency and diagnostic quality, providing a new reference direction for clinical practice.
