Streamlining Radiology Reporting: A hands-free approach with voice-to-text and generative AI

Purpose: In radiology, the generation of reports is critical for patient management, though can be time-consuming and labor-intensive. Studies have demonstrated the potential of AI in generating comprehensible and clinically relevant radiology reports. This study aims to take this concept further by investigating the feasibility of a hands-free radiology reporting system. This system would leverage voice-to-text technology and generative AI to streamline the reporting process. This approach has the potential to improve time efficiency, report standardization, and reduce radiologist workload.

Methods/Materials: This study utilized a prospective design. 100 chest X-ray reports were generated using both traditional methods and an automated process using generative AI. A radiologist dictated their findings into a voice-to-text software (Dragon) which created a raw text transcript. GPT-4 was then used to develop an AI-generated report based on our institution’s template. The radiologist’s traditional reports were used as comparisons, and the time to complete each voice-to-text transcript and traditional report was recorded. Two radiologists evaluated all reports for comprehensiveness, clarity, accuracy, clinical relevance, and conciseness via a 5 point Likert scale.

Results: The AI-driven system demonstrated considerable proficiency in accurately transcribing and generating radiology reports, albeit with occasional inaccuracies. While it achieved a high level of precision in standard scenarios, its performance varied in more complex cases. Most notably, the system substantially reduced the time taken to generate reports, signifying a potential increase in overall radiological workflow efficiency.

Conclusions: This study highlights the potential of voice-to-text and generative AI integration in radiology reporting. Increased time efficiency, even in standard scenarios, can allow radiologists to focus more on complex cases and patient care. However, further development and refinement are required to address comprehensiveness and occasional inaccuracies. Ongoing advancements in AI and speech recognition technologies suggest a positive outlook for future implementations in medical reporting.