Generative pretrained transformer-4, an artificial intelligence text predictive model, has a high capability for passing novel written radiology exam questions.

PURPOSE: AI-image interpretation, through convolutional neural networks, shows increasing capability within radiology. These models have achieved impressive performance in specific tasks within controlled settings, but possess inherent limitations, such as the inability to consider clinical context. We assess the ability of large language models (LLMs) within the context of radiology specialty exams to determine whether they can evaluate relevant clinical information. METHODS: A database of questions was created with official sample, author written, and textbook questions based on the Royal College of Radiology (United Kingdom) FRCR 2A and American Board of Radiology (ABR) Certifying examinations. The questions were input into the Generative Pretrained Transformer (GPT) versions 3 and 4, with prompting to answer the questions. RESULTS: One thousand seventy-two questions were evaluated by GPT-3 and GPT-4. 495 (46.2%) were for the FRCR 2A and 577 (53.8%) were for the ABR exam. There were 890 single best answers (SBA), and 182 true/false questions. GPT-4 was correct in 629/890 (70.7%) SBA and 151/182 (83.0%) true/false questions. There was no degradation on author written questions. GPT-4 performed significantly better than GPT-3 which selected the correct answer in 282/890 (31.7%) SBA and 111/182 (61.0%) true/false questions. Performance of GPT-4 was similar across both examinations for all categories of question. CONCLUSION: The newest generation of LLMs, GPT-4, demonstrates high capability in answering radiology exam questions. It shows marked improvement from GPT-3, suggesting further improvements in accuracy are possible. Further research is needed to explore the clinical applicability of these AI models in real-world settings.

Robotics in neurointerventional surgery: a systematic review of the literature.

BACKGROUND: Robotically performed neurointerventional surgery has the potential to reduce occupational hazards to staff, perform intervention with greater precision, and could be a viable solution for teleoperated neurointerventional procedures. OBJECTIVE: To determine the indication, robotic systems used, efficacy, safety, and the degree of manual assistance required for robotically performed neurointervention. METHODS: We conducted a systematic review of the literature up to, and including, articles published on April 12, 2021. Medline, PubMed, Embase, and Cochrane register databases were searched using medical subject heading terms to identify reports of robotically performed neurointervention, including diagnostic cerebral angiography and carotid artery intervention. RESULTS: A total of 8 articles treating 81 patients were included. Only one case report used a robotic system for intracranial intervention, the remaining indications being cerebral angiography and carotid artery intervention. Only one study performed a comparison of robotic and manual procedures. Across all studies, the technical success rate was 96% and the clinical success rate was 100%. All cases required a degree of manual assistance. No studies had clearly defined patient selection criteria, reference standards, or index tests, preventing meaningful statistical analysis. CONCLUSIONS: Given the clinical success, it is plausible that robotically performed neurointerventional procedures will eventually benefit patients and reduce occupational hazards for staff; however, there is no high-level efficacy and safety evidence to support this assertion. Limitations of current robotic systems and the challenges that must be overcome to realize the potential for remote teleoperated neurointervention require further investigation.