Natural Language Processing for Cardiology: A Narrative Review
Kailai Yang, Yan Leng, Xin Zhang, Tianlin Zhang, Paul Thompson, Bernard Keavney, Maciej Tomaszewski, Sophia Ananiadou · 2025
Read Natural Language Processing for Cardiology: A Narrative Review by Kailai Yang, Yan Leng, Xin Zhang, Tianlin Zhang, Paul Thompson, Bernard Keavney, Maciej Tomaszewski, Sophia Ananiadou (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-16708.txt.
Introduction
Kailai Yang and seven co-authors review how natural language processing has been used in cardiology from 2014 to 2025. Heart disease draws on many tangled factors, from genes to lifestyle to clinical history. Much of that information sits in free text, such as patient stories, medical records, and research papers, where standard statistics cannot reach it directly.
Searching six literature databases, the team screened the work down to two hundred sixty-five relevant articles. Each study was coded across several dimensions, covering diseases, data types, and methods. The review maps where language tools already help diagnosis, treatment, and prevention in heart care. It also shows where evidence remains thin for future work.
Catalog record
| CREATOR | Kailai Yang, Yan Leng, Xin Zhang, Tianlin Zhang, Paul Thompson, Bernard Keavney, Maciej Tomaszewski, Sophia Ananiadou |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 76 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-16708
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Natural Language Processing for Cardiology: A Narrative Review Kailai Yang, Yan Leng, Xin Zhang, Tianlin Zhang, Paul Thompson, Bernard Keavney, Maciej Tomaszewski, Sophia Ananiadou Abstract Cardiovascular diseases are becoming increasingly prevalent in modern society, with a profound impact on global health and well-being. These Cardiovascular disorders are complex and multifactorial, influenced by genetic predispositions, lifestyle choices, and diverse socioeconomic and clinical factors. Information about these interrelated factors is dispersed across multiple types of textual data, including patient narratives, medical records, and scientific literature. Natural language processing (NLP) has emerged as a powerful approach for analysing such unstructured data, enabling healthcare professionals and researchers to gain deeper insights that may transform the diagnosis, treatment, and prevention of cardiac disorders. This review provides a comprehensive overview of NLP research in cardiology from 2014 to 2025. We systematically searched six literature databases for studies describing NLP applications across a range of cardiovascular diseases. After a rigorous screening process, we identified 265 relevant articles. Each study was analysed across multiple dimensions, including NLP paradigms, cardiology-related tasks, disease types, and data sources. Our findings reveal substantial diversity within these dimensions, reflecting the breadth and evolution of NLP research in cardiology. A temporal analysis further highlights methodological trends, showing a progression from rule-based systems to large language models. Finally, we discuss key challenges and future directions, such as developing interpretable LLMs and integrating multimodal data. To the best of our knowledge, this review represents the most comprehensive synthesis of NLP research in cardiology to date. keywords Natural language processing, Cardiology, Clinical applications, Large language models † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. 1 Introduction Cardiovascular diseases (CVDs) comprise a group of disorders that affect the heart and blood vessels, posing a major global health challenge. The most common clinical manifestations include coronary artery disease, heart failure, arrhythmias, valvular heart disease, and hypertension. These conditions can significantly diminish an individual’s quality of life and overall well-being, with their prevalence continuing to rise due to aging populations and lifestyle changes ( Roth et al. 2020 ) . According to the World Health Organization, CVDs are the leading cause of death worldwide, accounting for an estimated 17.9 million deaths each year ( World Health Organization 2021 ) . Consequently, early detection and effective management of CVDs are essential to improving patient outcomes and mitigating their burden on healthcare systems. There are many types of text data that contain valuable information about the cardiovascular system. These include electronic health records (EHRs), clinical notes, radiology reports, research papers, and patient-reported outcomes,etc. In recent years, natural language processing (NLP), a branch of artificial intelligence (AI), has emerged as a powerful tool for analyzing and managing large-scale textual data in the field of cardiology. With the help of NLP algorithms, healthcare professionals can gain valuable insights from vast amounts of unstructured data, thus providing the potential to revolutionize approaches to cardiac diagnosis, treatment, and prevention. NLP techniques have been successfully applied to various tasks in cardiology, including the extraction of important cardiac measurements from echocardiography reports ( Nath et al. 2016 ) , classifying the severity of coronary artery disease from clinical narratives ( Gao et al. 2019 ) , predicting heart failure readmissions based on discharge summaries ( Golas et al. 2018 ) , identifying risk factors for CVDs, and supporting clinical decision-making ( Chen et al. 2020 ) . The rapidly evolving nature of NLP research promises to contribute towards further enhancements to patient care, advances in clinical research, and the development of more personalized and effective cardiovascular interventions ( Ahmed et al. 2025 ; Ferreira Santos et al. 2025 ) . The active nature of NLP research within the cardiology domain has already given rise to some brief reviews and surveys. For example, Turchioe et al. 2022 focused on identifying articles that describe the application of NLP techniques to EHRs referring to cardiology issues during the period 2015-2020. The review revealed that rule-based methods were dominant, although a smaller number of approaches based on traditional machine learning were also identified. The rapidly evolving nature of the NLP field has led to more reviews on the emerging technique of large language models (LLMs) ( Gala and Makaryus 2023 ; Ferreira Santos et al. 2025 ; Boonstra et al. 2024 ) within the cardiology domain, which have rapidly grown to be a dominant paradigm of many NLP approaches. Their training process on vast amounts of textual data facilitates sophisticated language understanding and text generation capabilities, which have been successfully applied to carry out a wide variety of related tasks. Despite the utility of the above reviews, there is currently a lack of a survey that provides a comprehensive account of the evolution of NLP within the cardiology domain that examines the full range of NLP techniques that have been applied in the field of cardiology over an extended time period, and how these techniques have been applied to address a variety of cardiology-related applications concerning a broad set of cardiovascular diseases. In response, this review article provides a detailed overview of the NLP applications within the cardiology domain over the last decade. In contrast to Turchioe et al. 2022 , which only identified 37 NLP-related articles containing general terms relating to cardiology (i.e., cardiology , cardiac or cardiovascular ), we have conducted a extensive search that covers a range of more specific cardiovascular diseases. This allows us to consider the impact of NLP in greater depth across various branches of cardiology. Specifically, we cover the period 2014 - 2025, which overlaps with, but extend beyond, the time span covered by Turchioe et al. 2022 , leading to coverage of crucial emerging trends such as LLM-based methods and novel perspectives such as retrieval-augmented generation. Our extended search parameters have also resulted in the identification of a much larger 265 number of relevant studies compared to previous reviews, thus permitting an in-depth analysis of the impact of NLP across a range of cardiovascular sub-fields, such as heart failure ( Girouard et al. 2025 ) and ECG diagnosis ( Ansari et al. 2025 ) . The wide span of related works that we cover also allows us to comprehensively analyze the evolution of NLP methods in cardiology, providing a systematic landscape of advancing NLP techniques – from traditional machine learning to LLMs – and their influence on various cardiovascular sub-fields. In summary, this review is aimed at answering the following research questions: • What types of NLP paradigms have been employed to address various text-related tasks and applications in cardiology? • To what extent have NLP approaches been applied across different cardiovascular diseases and types of textual sources? • What is the evolving landscape and emerging trends of NLP approaches for cardiology? • What is the key challenges and most promising directions for future research for more effective NLP applications in cardiology? 2 Search Methodology 2.1 Search strategy In order to identify relevant articles, we conducted a comprehensive search using six literature search platforms (i.e, PubMed, Scopus, Web of Science, IEEE Xplore, ACM Digital Library, and DBLP computer science bibliography) to find articles written in English and published between July 2014 and