Computer vision applications in healthcare: a literature review augmented with natural language processing techniques

• Autor(a) principal: Constâncio, Alex Sebastião
• Data de Publicação: 2022
• Outros Autores: Carvalho, Deborah Ribeiro, Tsunoda, Denise Fukumi
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/32942
• topic: Sistemas de visão computacional; Aprendizado profundo; Diagnóstico; Saúde; Medicina.; Computer vision systems; Deep learning; Diagnosis; Health; Medicine.; Sistemas de visión por ordenador; Aprendizaje profundo; Diagnóstico; Salud; Medicina.

Computer vision systems (CVS) have received special attention from researchers for their high adaptability to various contexts, especially in the security area for image and video recognition. This paper presents a literature review on the use of computer vision in healthcare over the past five years (2017-2021) and presents trends and analysis for the first six months of 2022. The Science Direct, Scopus, Web of Science, ACM Digital Library, and IEEE Xplore databases were used to conduct the search. A total of 2,072 articles were retrieved (2017 to 2021) and 492 articles in 2022 and of these, after deduplication, 1,857 papers composed the 2017-2021 corpus and 465 the 2022 corpus. Biblioshiny features (R's Bibliometrix package) were used for metrics such as journals that most publish on the topic and Natural Language Processing techniques were adopted to extract multigrams that generated word clouds from the abstracts of the retrieved articles. Brazil appears in only three papers: one by researchers from the Federal University of Acre, one from the State University of Maringa, and another from the Federal University of Santa Catarina, and all three are literature reviews. Chinese researchers appear as the most productive in the field and deep learning is the main technology adopted for this kind of study. The diseases most evidently explored in the period are breast cancer and COVID-19.