A Survey of Deep Learning Techniques and Applications in Bioengineering: A Latin American Perspective

The rapid advancement of new paradigms in artificial intelligence has had a profound impact on scientific research in recent years. The field of medical data and its applications in healthcare have particularly benefited from these advancements, rising the quality of diagnoses, treatment effectiveness, and knowledge extraction. Deep learning has revolutionized the concept of data, including images, videos, text, and signals. Convolutional neural networks and sequential models have played a significant role in feature extraction and classification of biomedical images and signals, enabling powerful diagnostic assistance. Generative models have further expanded the artificial intelligence capabilities, allowing the generation of realistic images and sequences. Attention mechanisms and Transformers have revolutionized the field, pushing the boundaries of the applications. Despite these advancements, accessibility to these technologies remains a challenge, especially in less developed countries where computational resources are limited. This paper provides a review of deep-learning applications in Bioengineering, focusing on the works of researchers from Latin America. It explores the types of applications and technologies during the period 2021 to 2023, which witnessed unprecedented advancements. Even with these limitations, there has been significant and high-quality production and progresses. However, efforts are required to ensure these technologies can benefit researchers and practitioners worldwide.