Comparative analysis of artificial intelligence tools for the development of medical imaging diagnostic platforms for respiratory diseases.

Abstract

Respiratory diseases caused by viruses and bacteria, such as COVID-19, bacterial pneumonia, and tuberculosis, pose a threat to public health in general, especially in contexts where access to high-tech medical tools is limited. In this sense, artificial intelligence has emerged as a key tool for optimizing medical imaging diagnosis, improving accuracy and reducing care times.The overall objective of this research was to identify, characterize, and compare technological tools that facilitate the development of AI-powered respiratory disease diagnostic platforms. A descriptive methodology with a qualitative approach was used, structured in four stages: document review, tool classification, design of a comparative matrix, and qualitative evaluation, generating a critical selection path.The results allowed us to identify and prioritize tools such as Python, TensorFlow, ChestX-ray8, ONNX, Pydicom, and Flask/FastAPI, which showed advantages in terms of performance, compatibility, scalability, and accessibility. These tools offer a solid foundation for the development of automated diagnostic platforms, particularly in resource-limited clinical settings.
This research provides validated and consistent guidance for the selection of technological tools in medical solutions projects, contributing to the development of robust, scalable, and ethically responsible clinical software.