Trustworthy AI for signals and image processing: a telemedicine perspective

Artficial Intelligence is showing unprecedented performance in signals & image processing. Classification, segmentation and generative process seem to have unlimited potential. The roots of Artificial Intelligence are deep in scientific history, but in the world of Big Data and Internet 5.0, its use and effects have yet to be entirely tested. The black box problem, security, privacy issues, and public opinion are some of the factors that push towards the development of a new concept: "Trustworthy AI". The use of advanced methods, such as EfficientNet & GradCAM, leads to remarkable accuracy and consistent explanation in the classification of ultrasound. Further studies aim at analyzing results could lead to a more robust application of AI in the generalized field of signal and image processing and will lay the foundation for future work on reliable AI.