Desenvolvimento de uma ferramenta automatizada para a geração de laudos médicos utilizando a API do GPT

Abstract

Machine Learning is a subfield of AI focused on the development of algorithms and models that enable computers to learn from experience, making predictions or decisions without being explicitly programmed for specific tasks. It plays an essential role in Generative AI, allowing models, such as GPT, to be trained with large volumes of data to generate text coherently and accurately. The ability to learn and replicate complex patterns makes these models ideal for applications requiring a high degree of precision and consistency, such as generating medical reports. With advancements in imaging diagnostic technologies (such as X-ray, Ultrasound, CT, and MRI), the number of imaging exams per patient has significantly increased, leading to an overload of professionals responsible for analyzing the images and preparing reports. This work aimed to develop a tool to automate the report generation process, so professionals can focus solely on the analysis of the imaging exams. The tool was developed using the Backend For Frontend architecture. The Backend was implemented in the Python programming language and is responsible for calling the GPT API to generate the reports and converting audio to text, should the physician choose to provide input via audio. The Frontend handles the user interface and interaction, running on a web browser accessible from any device, whether Desktop or Mobile, and was developed with HTML5, CSS3, and JavaScript. For testing purposes, a link to access the platform was provided to the radiologist. This allowed the professional the freedom to use, analyze, and evaluate the system’s functionality.[resumo fornecido pelo autor]