In an experiment, physician-researchers at Beth Israel Deaconess Medical Center (BIDMC) based in Boston conducted a study using Open AI's Chat-GPT 4 to assess its ability to accurately diagnose complex medical cases. The findings showed that Chat-GPT 4 successfully provided the correct diagnosis in almost 40% of cases and listed the correct diagnosis in two-thirds of challenging cases among its potential diagnoses. These promising results highlight the potential of AI-driven technology like Chat-GPT 4 in assisting medical professionals with accurate and insightful diagnostic capabilities.
Adam Rodman, co-director of the Innovations in Media and Education Delivery (iMED) Initiative at BIDMC, commented on the study's motivation, stating that recent advancements in artificial intelligence have produced generative AI models with highly proficient text-based responses, particularly in standardized medical examinations. The researchers sought to assess if such a generative model could "think" like a doctor and effectively solve standardized complex diagnostic cases used for educational purposes. The performance of Chat-GPT 4 exceeded expectations, providing promising insights for its potential applications in the medical field.
ALSO READ: iQOO Neo 7 Pro 5G Review: Powerful smartphone under Rs. 40,000
To evaluate the chatbot's diagnostic abilities, the researchers employed clinicopathological case conferences (CPCs), which involve complex patient cases with relevant clinical and laboratory data, imaging studies, and histopathological findings for educational purposes. Out of 70 CPC cases, the AI correctly matched the final CPC diagnosis in 39% of cases. Moreover, the AI's differential – a list of possible conditions based on a patient's symptoms, medical history, clinical findings, and test results – included the final CPC diagnosis in 64% of cases.
Zahir Kanjee, the study's first author and a hospitalist at BIDMC, emphasized that while chatbots cannot replace the expertise of trained medical professionals, generative AI shows promise as a potential adjunct to human cognition in diagnosis. It has the potential to assist physicians in understanding complex medical data and refining diagnostic thinking.
ALSO READ: OpenAI's chatbot ChatGPT now available on Android in India
The study highlights the promising capabilities of AI technology in the medical field. However, researchers acknowledge the need for further exploration of optimal uses, benefits, and limitations of AI models, especially concerning privacy issues. Understanding how these new AI technologies might transform healthcare delivery is crucial for their successful integration into medical practices.
Inputs from IANS
