The SARS-CoV-2 pandemic sparked unprecedented research efforts across multiple fields. Natural language processing (NLP) offers methods for analyzing scientific articles and represents a promising remedy to information overload in biomedical research. As a case study, we introduce an NLP-based solution to streamline the navigation and interpretation of literature on AI-based medical imaging of COVID-19. This field saw a significant increase in publications during the health emergency. Our solution features an interactive maplike dashboard that enables fine-grained visualization and summarization of a large corpus of articles. We evaluated various biomedical transformer models to extract input vectorial representations of each abstract in our corpus. By providing insights into topical and bibliometric patterns, this approach supports users’ orientation within the literature. Moreover, it facilitates knowledge discovery by covering a broader range of papers than traditional systematic reviews.