RESUMO
OBJECTIVES: Large language models (LLMs) have shown potential in radiology, but their ability to aid radiologists in interpreting imaging studies remains unexplored. We investigated the effects of a state-of-the-art LLM (GPT-4) on the radiologists' diagnostic workflow. MATERIALS AND METHODS: In this retrospective study, six radiologists of different experience levels read 40 selected radiographic [n = 10], CT [n = 10], MRI [n = 10], and angiographic [n = 10] studies unassisted (session one) and assisted by GPT-4 (session two). Each imaging study was presented with demographic data, the chief complaint, and associated symptoms, and diagnoses were registered using an online survey tool. The impact of Artificial Intelligence (AI) on diagnostic accuracy, confidence, user experience, input prompts, and generated responses was assessed. False information was registered. Linear mixed-effect models were used to quantify the factors (fixed: experience, modality, AI assistance; random: radiologist) influencing diagnostic accuracy and confidence. RESULTS: When assessing if the correct diagnosis was among the top-3 differential diagnoses, diagnostic accuracy improved slightly from 181/240 (75.4%, unassisted) to 188/240 (78.3%, AI-assisted). Similar improvements were found when only the top differential diagnosis was considered. AI assistance was used in 77.5% of the readings. Three hundred nine prompts were generated, primarily involving differential diagnoses (59.1%) and imaging features of specific conditions (27.5%). Diagnostic confidence was significantly higher when readings were AI-assisted (p > 0.001). Twenty-three responses (7.4%) were classified as hallucinations, while two (0.6%) were misinterpretations. CONCLUSION: Integrating GPT-4 in the diagnostic process improved diagnostic accuracy slightly and diagnostic confidence significantly. Potentially harmful hallucinations and misinterpretations call for caution and highlight the need for further safeguarding measures. CLINICAL RELEVANCE STATEMENT: Using GPT-4 as a virtual assistant when reading images made six radiologists of different experience levels feel more confident and provide more accurate diagnoses; yet, GPT-4 gave factually incorrect and potentially harmful information in 7.4% of its responses.
RESUMO
The objective of this randomized controlled trial (RCT) was to assess the impact of rifaximin on the course of liver function, liver regeneration and volumetric recovery in patients undergoing major hepatectomy. The ARROW trial was an investigator initiated, single-center, open-label, phase 3 RCT with two parallel treatment groups, conducted at our hepatobiliary center from 03/2016 to 07/2020. Patients undergoing major hepatectomy were eligible and randomly assigned 1:1 to receive oral rifaximin (550 mg twice daily for 7-10 or 14-21 days in case of portal vein embolization preoperatively and 7 days postoperatively) versus no intervention. Primary endpoint was the relative increase in postoperative liver function measured by LiMAx from postoperative day (POD) 4 to 7. Secondary endpoint were the course of liver function and liver volume during the study period as well as postoperative morbidity and mortality. Between 2016 and 2020, 45 patients were randomized and 35 patients (16 individuals in the rifaximin and 19 individuals in the control group) were eligible for per-protocol analysis. The study was prematurely terminated following interim analysis, due to the unlikelihood of reaching a significant primary endpoint. The median relative increase in liver function from POD 4 to POD 7 was 27% in the rifaximin group and 41% in the control group (p = 0.399). Further, no significant difference was found in terms of any other endpoints of functional liver- and volume regeneration or perioperative surgical complications following the application of rifaximin versus no intervention. Perioperative application of rifaximin has no effect on functional or volumetric regeneration after major hepatectomy (NCT02555293; EudraCT 2013-004644-28).
