Editor-DN
what a story
Chat GPTAn AI model based on OpenAI’s GPT-4 LLM has shown some impressive skills in diagnosing brain tumors.
Researchers at Osaka Metropolitan University’s Graduate School of Medicine conducted a study comparing the diagnostic accuracy of ChatGPT and radiologists using 150 preoperative brain tumor MRI reports.
The results were quite surprising: ChatGPT scored 73%, beating the accuracy of neuroradiologists (72%) and general radiologists (68%).
ChatGPT performance varies based on report authoring.
The study also found that the diagnostic accuracy of ChatGPT varied based on who wrote the clinical report.
When it was interpreting neurologists’ reports, the AI model’s accuracy reached 80%. But when it came to reports from general radiologists, that number dropped to 60 percent.
Therefore, it appears that the type of radiologist behind the report may indeed affect how accurate ChatGPT diagnoses are.
Future prospects of AI in clinical diagnosis
The research team, led by grad student Yasuhito Mitsuyama and associate professor Daiju Ueda, plans to further explore the potential of large language models such as ChatGPT in other diagnostic imaging fields.
Their purpose? Using AI to lighten the burden on doctors, increase diagnostic accuracy, and create a better educational environment.
“These results suggest that ChatGPT may be useful for preoperative MRI diagnosis of brain tumors,” Mitsuyama said.
