首页|Capital Medical University Reports Findings in Artificial Intelligence (Prelimin ary experiments on interpretable ChatGPT-assisted diagnosis for breast ultrasoun d radiologists)

Capital Medical University Reports Findings in Artificial Intelligence (Prelimin ary experiments on interpretable ChatGPT-assisted diagnosis for breast ultrasoun d radiologists)

扫码查看
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – New research on Artificial Intelligenc e is the subject of a report. According to news reporting originating in Beijing , People’s Republic of China, by NewsRx journalists, research stated, “Ultrasoun d is essential for detecting breast lesions. The American College of Radiology’s Breast Imaging Reporting and Data System (BI-RADS) classification system is wid ely used, but its subjectivity can lead to inconsistency in diagnostic outcomes. ” The news reporters obtained a quote from the research from Capital Medical Unive rsity, “Artificial intelligence (AI) models, such as ChatGPT-3.5, may potentiall y enhance diagnostic accuracy and efficiency in medical settings. This study aim ed to assess the utility of the ChatGPT-3.5 model in generating BI-RADS classifi cations for breast ultrasound reports and its ability to replicate the ‘chain of thought’ (CoT) in clinical decision-making to improve model interpretability. B reast ultrasound reports were collected, and ChatGPT-3.5 was used to generate di agnoses and treatment plans. We evaluated GPT-4’s performance by comparing its g enerated reports to those from doctors with various levels of experience. We als o conducted a Turing test and a consistency analysis. To enhance the interpretab ility of the model, we applied the CoT method to deconstruct the decision-making chain of the GPT model. A total of 131 patients were evaluated, with 57 doctors participating in the experiment. ChatGPT-3.5 showed promising performance in st ructure and organization (S&O), professional terminology and expres sion (PTE), treatment recommendations (TR), and clarity and comprehensibility (C &C). However, improvements are needed in BI-RADS classification, ma lignancy diagnosis (MD), likelihood of being written by a physician (LWBP), and ultrasound doctor artificial intelligence acceptance (UDAIA). Turing test result s indicated that AI-generated reports convincingly resembled human-authored repo rts. Reproducibility experiments displayed consistent performance. Erroneous rep ort analysis revealed issues related to incorrect diagnosis, inconsistencies, an d overdiagnosis. The CoT investigation supports the potential of ChatGPT to repl icate the clinical decision-making process and offers insights into AI interpret ability.”

BeijingPeople’s Republic of ChinaAsi aArtificial IntelligenceEmerging TechnologiesMachine LearningTuring Test

2024

Robotics & Machine Learning Daily News

Robotics & Machine Learning Daily News

ISSN:
年,卷(期):2024.(Oct.2)