Global trends and emerging frontiers in robot-assisted vascular intervention: a comprehensive bibliometric analysis

Abstract

Robot-assisted vascular intervention may reduce occupational radiation exposure, improve procedural stability, and expand access to specialized endovascular care, but its global research landscape and developmental trajectory remain unclear. Publications related to robot-assisted vascular intervention were retrieved from the Web of Science Core Collection from January 1, 2006, to February 28, 2026. After screening according to predefined inclusion and exclusion criteria, 161 eligible articles were included. Bibliometric and visual analyses were performed using VOSviewer and CiteSpace to evaluate publication trends, countries, institutions, journals, authors, citation patterns, keyword co-occurrence, thematic clusters, and emerging hotspots. Annual publications increased markedly after 2020 and peaked in 2024. Reviews (29.8%) and observational clinical studies (26.7%) predominated, whereas randomized controlled trials remained scarce (1.9%). CorPath GRX was the most frequently reported robot-assisted vascular intervention platform, while the emergence of ETcath200 and others newer systems reflected increasing global platform diversification. The United States was the leading contributor, followed by China and the United Kingdom. The University of California System, Houston Methodist, and the University of Toronto were the most productive institutions. Highly cited studies were dominated by the studies of robotic-assisted percutaneous coronary intervention, although peripheral and neurointerventional applications have become increasingly prominent. Keyword analyses suggested an evolution from early feasibility, safety, and radiation-protection concerns toward broader clinical translation, remote intervention, artificial intelligence, and autonomous navigation. Robot-assisted vascular intervention is a rapidly expanding, multidisciplinary field, but its evidence base remains relatively immature. Future research should emphasize multicenter validation, cross-specialty collaboration, procedural standardization, and intelligent robotic assistance to support safer and more reproducible endovascular care.

Introduction

Vascular diseases, mainly including coronary artery disease and stroke, remain major contributors to morbidity, disability, and mortality worldwide [1, 2]. Percutaneous vascular interventions have become essential minimally invasive treatments for these conditions [3, 4]. However, conventional manual procedures expose operators to cumulative radiation and ergonomic stress, and procedural performance remains highly dependent on operator experience [5, 6]. In addition, unequal access to specialized endovascular expertise remains a major challenge. Remote-controlled robotic endovascular platforms have therefore emerged as a promising strategy to reduce occupational hazards, improve procedural stability, and potentially expand access to specialized vascular care [7].

Robot-assisted vascular intervention enables device manipulation from a shielded or remote workstation and may provide more stable and reproducible control of guidewires, catheters, and other endovascular devices [8]. Early clinical research was mainly concentrated in robotic percutaneous coronary intervention (PCI), whereas more recent studies have extended to neurovascular, peripheral, aortic, and venous applications [8–10]. However, the current literature remains heterogeneous in vascular territory, robotic platform, procedural indication, study design, and endpoint definition, ranging from case reports and early clinical experience to observational studies, registries, and randomized controlled trials (RCTs) [11–13]. Given this heterogeneity and the multidisciplinary nature of the field, a comprehensive understanding of its research landscape, leading contributors, thematic evolution, and emerging directions remains limited [14].

As a complement to traditional reviews and meta-analyses, bibliometric analysis provides a quantitative approach to mapping publication trends, citation patterns, collaboration networks, and keyword evolution, which is particularly useful for clarifying both the scientific foundation and the direction of future innovation [15, 16]. Therefore, this study aimed to conduct a comprehensive bibliometric and visual analysis of robot-assisted vascular intervention to map the global research landscape, intellectual structure, and emerging frontiers.

Materials and methods

Data source and search strategy

All bibliographic data were retrieved from the Web of Science Core Collection (WoSCC) database of Clarivate Analytics. WoSCC covers high-quality journals and conference proceedings globally and provides comprehensive citation indexing services, making it one of the most widely used databases for high-quality bibliometric analysis and academic research trend evaluation. The search was performed on February 28, 2026, and the literature search covered the period from January 1, 2006, to February 28, 2026. The specific search strategy was as follows: TS= (“robotic*” OR “robot-assisted” OR “robotics” OR “master-slave” OR “CorPath” OR “Corindus” OR “Magellan” OR “Hansen Medical” OR “Robocath” OR “R-One” OR “Robotic Angioplasty”) AND TS= (“endovascular” OR “interventional” OR “vascular” OR “neurointerventional” OR “catheter*” OR “angiography” OR “stent*” OR “percutaneous coronary intervention” OR “PCI” OR “thrombectomy” OR “coil*” OR “embolization”) AND PY= (2006–2026).

Inclusion and exclusion criteria

To ensure the quality of the included literature and its high relevance to the topic, strict inclusion and exclusion criteria were established for this study.

Inclusion criteria:

1. Document types were restricted to original research articles or reviews;

2. The publication language was limited to English;

3. Studies were included if they involved robot-assisted endovascular intervention in clinical settings, animal or vascular model experiments, simulation studies with clear medical applications, or clinically oriented system evaluations.

Exclusion criteria:

1. Non-full-text publication types, such as meeting abstracts, editorials, letters, corrections, and book chapters;

2. Purely technical or engineering studies without clinical or medically relevant applications;

3. Non-endovascular procedures, such as Da Vinci robot-assisted open or laparoscopic surgeries (e.g., robot-assisted nephrectomy, inferior vena cava thrombectomy, and prostatectomy);

4. Non-target endovascular procedures, such as cardiac radiofrequency ablation and hepatic or uterine artery embolization;

5. Records with missing information or insufficient data to confirm their relevance to this study.

Data extraction and processing

Two independent researchers screened the titles and abstracts of the retrieved records according to the predefined inclusion and exclusion criteria. Any discrepancies were resolved through discussion with a third senior researcher until consensus was reached. Eligible records were exported from the Web of Science Core Collection as “Full Record and Cited References” in plain-text format. The extracted metadata included publication title, authors, institutional affiliations, source journal, publication year, document type, citation counts, cited references, author keywords, and Keywords Plus. Before bibliometric analysis, author names, institutional names, journal titles, and keywords were standardized using a semi-automated thesaurus-based process supplemented by manual verification to reduce entity fragmentation and improve the accuracy of network analyses.

In addition to bibliometric variables, each included publication was further classified according to study type and robotic platform. Study types were categorized as review articles, RCT, observational clinical studies, case reports/case series/early clinical experience, or preclinical/model/simulation/technical evaluation studies. Robotic platform names were manually standardized when minor spelling or capitalization differences were present. For platform-specific temporal analysis, we focused on clinical publications, including RCT, observational clinical studies, and case reports/case series/early clinical experience, because review articles and some technical studies were not always attributable to a single clinically used platform.

Analytical tools and visual analysis

This study primarily used two widely adopted bibliometric software packages—VOSviewer (version 1.6.20) and CiteSpace (version 6.4.R1)—for data mining and scientific knowledge mapping.

• VOSviewer: Developed by Leiden University in the Netherlands, this software excels at constructing and visualizing large bibliometric networks. In this study, VOSviewer was used to construct author citation and keyword networks. Additionally, through keyword co-occurrence analysis, clustering algorithms were employed to identify the core research directions in this field (e.g., coronary interventions, neurovascular directions, and technological evolution).

• CiteSpace: Developed by Professor Chaomei Chen, this software focuses on exploring the evolutionary dynamics and frontier hotspots of specific disciplines. In this study, CiteSpace was used to construct the keyword co-occurrence network and the timeline view of keyword clusters, perform keyword burst detection, identify emerging research hotspots, and trace the temporal evolution of major research themes in robot-assisted endovascular intervention. ‘Burstness’ refers to a sudden increase in the citation rate or publication volume of literature, countries/regions, journals, or authors. The time span was set from 2010 to 2026, with a one-year time slice; the selection criteria were set as g-index with k = 25.

Furthermore, annual publication trends and other descriptive figures were generated using Microsoft Excel.

Local citations (LCs) represent citations received from other publications within the analyzed dataset, whereas global citations (GCs) represent the total citations recorded in the Web of Science Core Collection. The LC/GC ratio reflects the relative influence of a publication within the field-specific dataset compared with its broader citation impact, while normalized citation values facilitate comparisons across publication years.

Results

A total of 6,230 records were retrieved from the Web of Science database. The literature identification and screening process is shown in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram (Fig. 1). After excluding records that were neither original articles nor reviews (n = 613) and non-English publications (n = 117), 5,500 records remained for title and abstract screening. Of these, 1,723 were excluded because they were purely technical or engineering studies, 2,321 because they involved non-endovascular procedures, 872 because they focused on non-target endovascular procedures, and 423 for other reasons unrelated to the scope of this study. Ultimately, 161 eligible publications were included in the bibliometric analysis.

Fig. 1.

PRISMA flow diagram of literature identification, screening, and inclusion

Among the 161 included publications, review articles accounted for the largest proportion (n = 48, 29.8%), followed by observational clinical studies (n = 43, 26.7%), preclinical/model/simulation/technical evaluation studies (n = 37, 23.0%), case reports/case series/early clinical experience (n = 30, 18.6%), and RCT (n = 3, 1.9%) (Table 1). The temporal distribution of study types showed that publications increased markedly after 2020, but evidence from RCT remained scarce throughout the study period (Fig. 2A).

Table 1.

Study-type distribution of the 161 included publications in robotic endovascular intervention

Study type	No.	%
Review articles	48	29.80%
RCT	3	1.90%
Observational clinical studies	43	26.70%
Case reports/case series/early clinical experience	30	18.60%
Preclinical/model/simulation/technical evaluation studies	37	23.00%
Total	161	100.00%

Fig. 2.

Temporal distribution of study types and platform-specific publication trends in robotic endovascular intervention. (A) Stacked bar chart showing the annual distribution of the 161 included publications by study type. (B) Bubble plot showing platform-specific temporal trends among clinical publications. Bubble size and color intensity indicate the number of publications for each robotic platform in a given year. RCT, randomized controlled trial

Platform-specific trends were further examined among clinical publications, including RCT, observational clinical studies, and case reports/case series/early clinical experience. CorPath GRX was the most frequently reported platform in clinical publications (n = 42), followed by CorPath 200 (n = 14), Magellan/Sensei (n = 7), other systems (n = 6), Robocath R-One (n = 4), and ETcath200 (n = 3). Temporally, CorPath 200 appeared mainly in earlier clinical publications between 2011 and 2020, whereas CorPath GRX emerged from 2018 onward and became the dominant platform in subsequent clinical reports, with the highest number of platform-specific clinical publications in 2024. Magellan/Sensei appeared intermittently between 2013 and 2020. In contrast, Robocath R-One and ETcath200 were mainly represented in more recent publications, suggesting ongoing diversification of robotic endovascular platforms (Fig. 2B).

Annual publication trend

Although the literature search covered 2006–2026, the first eligible publications appeared in 2010; therefore, annual publication trends were visualized from 2010 to 2026 (Fig. 3A). Before 2019, fewer than 10 articles were published annually. Publication output increased rapidly after 2020 and peaked in 2024 with 33 publications. Although the number decreased slightly in 2025, it remained higher than the pre-2020 level. The polynomial regression model using data from 2010 to 2025 showed a good fit (R² = 0.8359), and the projected number of publications for 2026 was approximately 34 [17].

Fig. 3.

Global publication trends and collaborative landscape of robot-assisted vascular intervention. (A) Annual publication output from 2010 to 2026 with polynomial fitted trend line; the 2026 value is shown separately because the year was incomplete at the time of data retrieval. (B) Country-level collaboration map based on author affiliation information; line thickness indicates the number of joint publications between country pairs. (C) Publication output by country, stratified by single-country and multicountry publications. (D) Top 10 most productive institutions in this field

Countries, institutions, authors and journals

Collaborative networks

Figure 3B shows the country-level co-authorship network based on author affiliation information. In this map, links represent joint publications involving authors from institutions located in different countries, rather than formal national-level collaborative programs. The United States had the most prominent presence in the network, followed by China and Canada. The strongest country-pair link was observed between the United States and Canada, with 6 joint publications, while the United States–Hungary link was also prominent, with 5 joint publications. These findings suggest that international collaboration in this field was present but concentrated in a limited number of country pairs.

Geographic and national contributions

Among the top 10 contributing countries, the United States ranked first with 66 publications, followed by China (22) and the United Kingdom (15) (Fig. 3C). The United States also had the highest numbers of both single-country publications (n = 48) and multiple-country publications (n = 18). Canada showed the highest proportion of multicountry co-authored publications (85.7%), whereas Japan and Brazil had no multicountry co-authored publications identified in this dataset based on author affiliation information.

Institution analysis

The institutional landscape revealed a high degree of research concentration, with a small number of academic centers leading the technological development of endovascular robot. The University of California System and Houston Methodist were the most productive institutions, each contributing 32 publications, followed by the University of Toronto with 25 publications (Fig. 3D). Together, these institutions formed the principal centers of research activity in this field.

Most productive authors

Table 2 shows the top 10 most productive authors in robot-assisted vascular intervention. Alan B. Lumsden ranked first with 13 publications, followed by Ehtisham Mahmud with 11. Vitor Mendes Pereira and Gavin W. Britz each contributed 9 publications. Lumsden and Pascal M. Jabbour had the highest H-index values, while Ehtisham Mahmud had the highest total citations. Most leading authors were affiliated with US institutions, with additional contributions from Canada and the UK.

Table 2.

Top 10 most productive authors

Rank	Author	Institution	No. of Publications	H-index	Total Number of Citations	Average Number of Citations
1	Lumsden, Alan B.	Houston Methodist Debakey Heart & Vasc Ctr, FanninSt, Houston, TX USA	13	65	249	19.15
2	Mahmud, Ehtisham	Univ Calif San Diego, Div Cardiovasc Med, La Jolla, CA 92,093 USA	11	38	433	39.36
3	Pereira, Vitor Mendes	Univ Toronto, St Michaels Hosp, Neurovasc Ctr, Toronto, ON, Canada Univ Toronto, St Michaels Hosp, RADIS Lab, Toronto, ON, Canada	9	18	205	22.78
4	Britz, Gavin W.	Methodist Neurol Inst, Dept Neurosurg, Houston, TX USA	9	42	153	17
5	Riga, Celia V.	St Marys Hosp, Imperial Coll Healthcare, Reg Vasc Unit, Dept Vasc Surg & Intervent Radiol, London W2 1NY, England Univ London Imperial Coll Sci Technol & Med, Dept Surg & Canc, London, England	8	32	216	27
6	Ang, Lawrence	Univ Calif San Diego, Div Cardiovasc Med, La Jolla, CA 92,093 USA	7	13	284	40.57
7	Cancelliere, Nicole Mariantonia	St Michaels Hosp, RADIS Lab, Toronto, ON, Canada	7	22	186	26.57
8	Reeves, Ryan R.	Univ Calif San Diego, Sulpizio Cardiovasc Ctr, La Jolla, CA USA	6	20	289	48.17
9	Sungur, John-Michael	Corindus, Siemens Healthineers Co, Waltham, MA USA	6	6	207	34.5
10	Jabbour, Pascal M.	Thomas Jefferson Univ Hosp, Dept Neurol Surg, Philadelphia, PA USA	6	65	133	22.17

Journal analysis

The H-index measures academic impact by counting the number of papers with at least h citations each, reflecting a researcher’s or journal’s contribution. Catheterization and Cardiovascular Interventions published the largest number of articles (n = 15, 9.3%), followed by the Journal of NeuroInterventional Surgery (n = 10, 6.2%) and Interventional Neuroradiology (n = 9, 5.6%) (Table 3). The top 10 journals published 71 articles, accounting for 44.1% of all included publications; eight of these journals were ranked in JCR Q1 or Q2.

Table 3.

Top 10 most prolific journals

Rank	Prolific Journal	Count	Percentage of all publications	JCR	H-Index
1	Catheterization and Cardiovascular Interventions	15	9.317%	Q3	134
2	Journal of Neurointerventional Surgery	10	6.211%	Q1	75
3	Interventional Neuroradiology	9	5.590%	Q2	30
4	Journal of Invasive Cardiology	8	4.969%	Q3	66
5	International Journal of Computer Assisted Radiology and Surgery	6	3.727%	Q2	37
6	International Journal of Medical Robotics and Computer Assisted Surgery	6	3.727%	Q2	44
7	Journal of Vascular Surgery	5	3.106%	Q1	178
8	Cardiovascular and Interventional Radiology	4	2.484%	Q2	99
9	Journal of Geriatric Cardiology	4	2.484%	Q2	21
10	Journal of Neurosurgery	4	2.484%	Q1	189

Among the top 10 cited journals, the Journal of the American College of Cardiology and JACC: Cardiovascular Interventions ranked first and second, with 228 and 210 citations, respectively (Table 4). Together, the JACC series accounted for 32% of citations among the top 10 cited journals. Other highly cited sources included Catheterization and Cardiovascular Interventions, EuroIntervention, Circulation, and the Journal of NeuroInterventional Surgery.

Table 4.

Top 10 most cited journals

Rank	Cited journals	Frequency	Percentage of citations among the top 10 cited journals	JCR	H-Index
1	Journal of the American College of Cardiology	228	17%	Q1	394
2	JACC: Cardiovascular Interventions	210	15%	Q1	163
3	Catheterization and Cardiovascular Interventions	186	14%	Q3	134
4	EuroIntervention	177	13%	Q1	101
5	Circulation	141	10%	Q1	570
6	Journal of NeuroInterventional Surgery	100	7%	Q1	75
7	Journal of Invasive Cardiology	86	6%	Q3	66
8	Journal of Vascular Surgery	81	6%	Q1	178
9	Cardiovascular Revascularization Medicine	80	6%	Q3	47
10	International Journal of Medical Robotics and Computer Assisted Surgery	77	6%	Q2	44

Article analysis

The 10 most locally cited articles are presented in Table 5. The PRECISE study [11], published in 2013, ranked first in local citations (LC = 84; GC = 254), followed by the CORA-PCI study [8] in 2017 (LC = 52; GC = 137). The 2019 first-in-human report of long-distance tele-robotic PCI [18] had the highest normalized LC value (normalized LC = 4.46). In addition to robotic PCI studies, the top-cited list included robotic peripheral vascular intervention and robotic neuroendovascular intervention studies, indicating increasing attention to non-coronary vascular applications [12, 19].

Table 5.

Top 10 locally cited articles

Rank	Article Title	Journal	Publication year	LC	GC	LC/GC Ratio (%)	Normalized LC	Normalized GC
1	Safety and Feasibility of Robotic Percutaneous Coronary Intervention Precise (percutaneous Robotically-Enhanced Coronary Intervention) Study [11]	Journal of the American College of Cardiology	2013	84	254	33.07	1.71	1.67
2	Demonstration of the Safety and Feasibility of Robotically Assisted Percutaneous Coronary Intervention in Complex Coronary Lesions Results of the Cora-PCI Study (complex Robotically Assisted Percutaneous Coronary Intervention) [8]	JACC: Cardiovascular Interventions	2017	52	137	37.96	3.96	2.80
3	Long Distance Tele-Robotic-Assisted Percutaneous Coronary Intervention: a Report of First-In-Human Experience [18]	EclinicalMedicine	2019	39	123	31.71	4.46	4.35
4	First-In-Human Evaluation of a Novel Robotic-Assisted Coronary Angioplasty System [20]	JACC: Cardiovascular Interventions	2011	34	131	25.95	2.67	1.93
5	Feasibility and Safety of Robotic Peripheral Vascular Interventions Results of the RAPID Trial [12]	JACC -Cardiovascular Interventions	2016	34	80	42.50	3.29	2.49
6	Robotic-Enhanced Pci Compared to the Traditional Manual Approach [21]	Journal of Invasive Cardiology	2014	26	57	45.61	1.95	1.61
7	Safety and Feasibility of a Novel, Second-Generation Robotic-Assisted System for Percutaneous Coronary Intervention: First-In-Human Report [22]	Journal of Invasive Cardiology	2018	26	76	34.21	4.33	4.09
8	First-In-Human, Robotic-Assisted Neuroendovascular Intervention [19]	Journal of NeuroInterventional Surgery	2020	23	112	20.54	2.73	3.52
9	Longitudinal Geographic Miss (lgm) in Robotic Assisted Versus Manual Percutaneous Coronary Interventions [23]	Journal of Interventional Cardiology	2015	19	46	41.30	2.38	1.50
10	Comparison of Robotic Percutaneous Coronary Intervention with Traditional Percutaneous Coronary Intervention a Propensity Score-Matched Analysis of a Large Cohort [24]	Circulation-Cardiovascular Interventions	2020	19	58	32.76	2.25	1.82

Citation-based author network

The citation-based author network identified three major clusters (Fig. 4A). The green cluster, represented by Ehtisham Mahmud, Giora Weisz, and Ryan R. Reeves, was mainly associated with coronary interventional robotics. The red cluster, including Patrick Nicholson, John-Michael Sungur, Ivan Radovanovic, and Timo Krings, was related to neurointerventional robotics. The blue cluster, represented by Alan B. Lumsden, Jean Bismuth, Celia Riga, and Ponraj Chinnadurai, was associated with peripheral and broader endovascular robot applications. Gavin W. Britz occupied an intermediate position between the clusters.

Fig. 4.

Intellectual structure and keyword evolution of robot-assisted vascular intervention. (A) Citation-based author network showing citation-related intellectual communities among influential authors. (B) VOSviewer author keyword cluster analysis. (C) CiteSpace keyword co-occurrence network. (D) CiteSpace keyword cluster timeline showing the temporal evolution of major research themes

Keyword co-occurrence

Figure 4B presents the keyword clustering network generated by VOSviewer, illustrating the thematic structure of research on robot-assisted vascular intervention. The network suggests that this field is organized around several interconnected themes. The blue cluster, represented by keywords such as “percutaneous coronary intervention,” “coronary artery disease,” “radiation exposure,” and “robotics,” reflects the dominant role of coronary applications in the early development of vascular interventional robotics. The red cluster, including “stroke,” “aneurysm,” “brain,” “stent,” and “angiography,” indicates the extension of robotic systems into neurovascular intervention and other complex endovascular procedures. The green cluster, characterized by terms such as “medical robotics,” “robotic surgical procedures,” “endovascular procedures,” “neurointervention,” and “interventional radiology,” reflects broader clinical and procedural applications of robotic technology across vascular specialties. Meanwhile, the cyan cluster, including “artificial intelligence,” “machine learning,” and “interventional cardiology,” suggests an emerging research direction toward intelligent guidance, image-based navigation, and automation. Overall, the clustering network indicates that the field has evolved from feasibility and safety validation in coronary intervention toward broader vascular applications and intelligent robotic-assisted platforms.

Figure 4C shows the keyword co-occurrence network generated by CiteSpace, reflecting major research hotspots in robot-assisted vascular intervention. Larger nodes indicate higher keyword frequency, whereas nodes with purple outer rings indicate higher betweenness centrality. The most frequent keywords were “feasibility,” “safety,” and “percutaneous coronary intervention,” suggesting that current research has mainly focused on clinical feasibility, procedural safety, and coronary applications. Other keywords, including “radiation exposure,” “cardiologists,” “PCI,” and “coronary artery disease,” further highlight the importance of radiation protection and procedural performance. In addition, keywords such as “stroke,” “embolization,” “aneurysm repair,” “navigation,” “artificial intelligence,” and “machine learning” indicate the expansion of this field toward neurovascular intervention, peripheral vascular applications, and intelligent robotic technologies.

Keyword evolution and emerging trends

Figure 4D presents the keyword timeline view generated by CiteSpace, showing the temporal evolution of research topics in robot-assisted vascular intervention. The main clusters included “PVI,” “coronary artery disease,” “catheters,” “robotic surgical procedures,” “artificial intelligence,” “catheter navigation,” “aneurysm coiling,” “remote neuro-intervention,” “automatic medical systems,” and “chronic total occlusion.” Early keywords such as “safety,” “feasibility,” “angiography,” “system,” “radiation exposure,” and “percutaneous coronary intervention” indicate that initial studies mainly focused on technical feasibility, procedural safety, radiation protection, and coronary intervention. Over time, terms such as “catheter navigation,” “aneurysm coiling,” “PVI,” and “remote neuro-intervention” appeared, suggesting a gradual expansion toward broader endovascular and neurovascular applications. More recent keywords, including “artificial intelligence,” “robotic surgical procedures,” and “automatic medical systems,” indicate increasing attention to intelligent guidance, automation, and remote robotic intervention.

Figure 5 shows the top 20 keywords with the strongest citation bursts, further reflecting changes in research priorities over time. Early burst keywords, including “radiation” (strength = 1.53; 2010–2015), “vascular intervention,” “system,” and “hazards,” were mainly related to system development, procedural risk, and occupational radiation exposure. Subsequently, keywords such as “navigation,” “aneurysm repair,” “impact,” “percutaneous coronary intervention,” “safety” (strength = 2.90; 2020–2021), and “robotic surgical procedures” reflected increasing attention to clinical validation and endovascular applications. In the most recent period, “coronary artery disease” (strength = 2.52; 2025–2026) and “artificial intelligence” (strength = 2.35; 2025–2026) emerged as active burst keywords, indicating current interest in disease-specific robotic applications and AI-assisted navigation, automation, and decision support.

Fig. 5.

The top 20 keywords with the strongest citation bursts

Discussion

This study provides a comprehensive bibliometric overview of the global research landscape, intellectual structure, and emerging frontiers of robot-assisted vascular intervention. Several key findings are noteworthy. First, the marked increase in publications after 2020 likely reflects the transition of endovascular robot systems from limited early experience to more active clinical evaluation. Second, although the United States remains the leading contributor and China has emerged as an important participant, international collaboration remains uneven and appears to be concentrated around a limited number of countries and research groups. Third, robot-assisted PCI remains the most mature clinical application in the field, whereas peripheral, aortic, and neurovascular applications are increasingly shaping its next stage of development. Finally, the recent emergence of artificial intelligence (AI) and automation-related keywords indicates that the field is no longer focused solely on remote mechanical manipulation, but is beginning to explore more intelligent and reproducible models of endovascular care.

The distribution of study types indicates that robot-assisted endovascular intervention remains a relatively young and evolving field. Over the past decade, the literature has been dominated by reviews, animal and model-based studies, technical evaluations, early clinical experience, case reports, and observational studies, indicating that the field remains focused primarily on feasibility, safety, and technical validation. Meanwhile, continuous advances in robotic platforms, software control, and device compatibility suggest that the field is still undergoing active innovation and expansion. However, RCTs remain limited, indicating that the rapid increase in publications should not be equated with a mature clinical evidence base. Further prospective, multicenter, and comparative studies are needed to establish the clinical value of robotic endovascular intervention.

Platform-specific trends may reflect technological iteration and progressive expansion of clinical indications. Early clinical studies primarily evaluated the CorPath 200 and Magellan/Sensei systems. CorPath 200, the first-generation CorPath system, received clearance from the U.S. Food and Drug Administration in 2012 for robot-assisted PCI and was initially evaluated in coronary intervention [11]. In contrast, Magellan was developed to facilitate navigation within the peripheral vasculature and support manual delivery of therapeutic devices, representing an early extension of robotic assistance beyond coronary intervention [14]. The subsequent predominance of CorPath GRX likely reflects the transition to a second-generation platform with guide catheter control, improved precision, optimized workflow, and broader procedural capability. More recent systems, such as Robocath R-One, ETcath200, PANVIS-A, and YDHB-NS01, indicate increasing platform diversification, with efforts to improve device compatibility, remote manipulation, radiation protection, and adaptability to complex vascular territories. Taken together, these platform-level changes suggest that robotic endovascular intervention has evolved from early navigation-oriented or coronary-focused systems toward more versatile, territory-adapted platforms. However, because this analysis is based on publication records, the observed platform trajectories should be interpreted as trends in the published evidence rather than direct evidence of real-world adoption, procedural volume, or discontinuation of specific systems.

The annual publication trend suggests a clear expansion of research activity in robot-assisted vascular intervention. Early publication activity was relatively limited, reflecting the initial stage of robotic platform development and restricted clinical adoption [8, 25]. The marked increase after 2020 was likely driven by the maturation of robot-assisted interventional technologies, increasing clinical familiarity, and broader investigation across hospitals and specialized centers [25, 26]. This trend suggests that the field is progressing from preliminary feasibility evaluation toward broader clinical validation, with recent studies increasingly incorporating comparative designs, multicenter cohorts, and randomized trials. The polynomial regression model further predicted continued growth in 2026, supporting the view that robot-assisted vascular intervention remains an active and expanding research area.

Geographically, the research landscape remains uneven despite growing internationalization. The United States dominated publication output and occupied a central position in the affiliation-based co-authorship network, with the USA–Canada link showing the greatest strength and the USA–Hungary link also being prominent. However, these links should be interpreted as country-level co-authorship patterns based on author affiliations rather than evidence of broad national-level collaboration. In a relatively small and specialized field, strong bilateral links may be driven by repeated publications from specific institutions, clinical trial groups, or recurring author networks [8, 11]. China ranked second in output, reflecting its rapidly expanding contribution [13, 27, 28]. Canada, despite a smaller publication volume, showed the highest proportion of multicountry co-authored publications in this dataset and appeared to play a bridging role in international collaboration [29]. By contrast, Japan and Brazil showed limited visible cross-country co-authorship. However, it should not be interpreted as evidence of purely local research activity, because industry collaboration, platform provision, and technical support may not be fully reflected in author affiliations [29–31]. Overall, these findings suggest that future progress in vascular interventional robotics will require stronger cross-national collaboration, shared clinical datasets, multicenter validation, and coordinated development of technical and procedural standards.

At the author level, research activity was concentrated among several leading clinical teams, mainly from the United States. Their work spanned different vascular territories, including coronary, neurovascular, and peripheral interventions, reflecting the multidisciplinary expansion of robotic vascular intervention. However, the concentration of productive authors within a few institutions also indicates the need for broader international collaboration.

The journal and author citation patterns further underscore the multidisciplinary but historically coronary-centered nature of the field. High-impact cardiovascular journals and robotic PCI-related author communities have contributed substantially to the intellectual foundation of the field, whereas neurointerventional and peripheral robotic applications have formed smaller but increasingly visible research communities [32]. The predominance of JCR Q1/Q2 journals suggests that vascular interventional robotics has gained visibility in influential academic journals, although further multicenter validation and methodological standardization remain necessary.

Three major citation-based research communities emerged from the author network. Because this network was constructed from citation relationships rather than co-authorship links, the clusters should be interpreted as groups of authors with related intellectual influence and thematic proximity, not necessarily as distinct collaboration groups. The green cluster was mainly associated with robotic PCI and early clinical validation, including feasibility, safety, complex lesion treatment and remote PCI [8, 11, 33]. The blue cluster focused on peripheral and broader endovascular robot applications, particularly robotic catheter manipulation, image-guided navigation, and complex peripheral or aortic procedures [34, 35]. The red cluster was associated with neurointerventional robotics, including anatomically complex and technically demanding procedures such as aneurysm treatment, cerebral angiography, and carotid or intracranial interventions [36–38]. Authors positioned between clusters, such as Gavin W. Britz, may serve as citation bridges linking coronary, peripheral, and neurovascular research themes [37, 39]. Overall, this network suggests a shift from coronary-centered validation toward broader endovascular applications, accompanied by increasing integration across cardiology, vascular surgery, neurointervention, imaging, and biomedical engineering.

Among the most-cited articles, the 2013 PRECISE study supported the safety and feasibility of robotic PCI using the CorPath 200 system and reported a clinical success rate of 98.8% and a reduction of up to 95% in operator radiation exposure [11]. Together with subsequent studies on complex coronary lesions [8], second-generation robotic systems [24], and remote robotic PCI, these studies helped establish coronary intervention as the initial clinical foundation of this field. Notably, the 2019 first-in-human report of long-distance tele-robotic PCI illustrated the feasibility of remote robotic intervention and provided an important proof of concept for expanding access to specialized interventional care across geographic barriers [18]. In addition, influential studies such as the RAPID trial in peripheral vascular intervention [12] and early reports of robotic neuroendovascular procedures [19] indicate that the field is gradually extending toward broader endovascular applications. Taken together, these highly cited studies reflect several major themes that have shaped the clinical literature on vascular interventional robotics, including coronary feasibility validation, remote robotic intervention, and exploratory applications in peripheral and neurovascular procedures.

Keyword trend analysis suggests a phased development of vascular interventional robotics. In the early stage (2010–2015), research was driven primarily by concerns regarding occupational radiation exposure and procedural safety, as reflected by the sustained appearance of keywords such as “radiation” and “safety.” These themes indicate that robotic systems were initially developed, in part, to reduce radiation exposure and ergonomic burden among interventional physicians by enabling remote manipulation away from the bedside [5, 6]. During the application expansion phase (2016–2020), the emergence of “coronary artery disease” reflected a shift from feasibility assessment toward clinical application, with robotic PCI becoming the earliest and most extensively validated scenario because of its relatively standardized workflow and established evidence base [8, 40]. Since 2021, keywords related to neurointervention, stroke, and remote intervention have become more prominent, suggesting growing interest in applying robotic systems to more technically demanding and time-sensitive procedures. In this context, robotic platforms may offer potential advantages in motion stability, tremor reduction, and remote access to specialized neurointerventional expertise, particularly in regions with limited specialist availability [7, 10, 40].

Future research should advance along three complementary directions. First, the growing attention to AI indicates a shift from mechanical teleoperation toward intelligent robotic assistance. Emerging work has explored image-guided navigation, motion planning, learning from demonstration, and reinforcement learning–based catheter control, but fully autonomous endovascular intervention remains largely experimental and has not entered routine clinical practice [41, 42]. These technologies may help reduce operator fatigue, standardize procedural workflows, and improve reproducibility. However, their clinical translation will require clear safety boundaries, reliable human–machine interaction, lower robotic system costs and robust failure-response mechanisms [43, 44].

Second, the field should move from feasibility-oriented reporting toward territory-specific comparative validation. Although coronary intervention currently has the most developed evidence base, clinical evidence for neurovascular applications remains limited, particularly for more complex procedures such as intracranial stenting and aneurysm embolization [45]. In addition, larger, prospective, multicenter RCTs are needed to clarify the clinical value of robot-assisted vascular intervention.

Third, future studies should place greater emphasis on patient-centered benefits. Existing research has largely focused on technical feasibility, procedural success, and operator radiation protection, whereas evidence regarding direct patient benefits remains insufficient. Future investigations should focus on reducing patient radiation exposure and contrast use, shortening procedure time, decreasing conversion-to-manual rates, and enhancing the overall patient experience. Such evidence will be essential for establishing robotic endovascular intervention not only as a radiation-sparing tool for operators but also as a clinically meaningful strategy for improving patient care.

Several limitations should be acknowledged. First, citation-related bias is inherent to bibliometric analysis, as older publications may be overrepresented and recent but potentially influential studies, particularly those on AI-assisted navigation, autonomous catheter control, and remote intervention, may be underestimated. Second, this analysis was based exclusively on the Web of Science Core Collection, so relevant publications indexed only in other databases may have been missed, potentially influencing publication counts, citation rankings, collaboration patterns, and network mapping. And the results may also be affected by author or institutional name disambiguation, citation practices, and visualization parameters. Accordingly, the findings should be interpreted as a WoSCC-based bibliometric overview. Third, country-level collaboration was inferred from author affiliation-based co-authorship records and may not represent broad national-level collaboration. Strong bilateral links may be driven by repeated publications from a small number of institutions or author groups, while industry involvement, platform supply, and technical support may not be fully captured by author affiliations. Fourth, citation-based author clusters were derived from citation relationships rather than direct co-authorship links. Therefore, they reflect citation-related intellectual proximity and thematic influence, but cannot determine whether authors within or between clusters directly collaborated. Finally, bibliometric analysis can characterize publication patterns and knowledge structures but cannot directly evaluate the clinical efficacy, safety, or technical superiority of robot-assisted vascular intervention. Therefore, future studies should combine bibliometric mapping with evidence synthesis and expert-based qualitative assessment.

Conclusion

In conclusion, this bibliometric analysis provides a comprehensive overview of the clinical and translational research landscape of robot-assisted vascular intervention. The field has shown rapid growth since 2020, with the United States, China, and several leading academic centers serving as major contributors. Current evidence and citation patterns indicate that robot-assisted PCI remains the most mature and influential application, while peripheral, aortic, and neurovascular interventions are emerging as important areas of expansion. Keyword and citation analyses further suggest a developmental trajectory from early concerns regarding radiation protection and procedural safety toward broader clinical translation, remote intervention, and intelligent robotic assistance. As vascular interventional robotics continues to evolve, future research should prioritize multicenter clinical validation, cross-specialty collaboration, technical standardization, and the integration of AI to support safer, more precise, and more reproducible endovascular care.

Author contributions

X.W., J.H., S.G., and X.Z. contributed to conceptualization, methodology, formal analysis, visualization, and writing of the original draft. Z.X., S.C., L.Z., X.L., Y.S., and J.L. contributed to data curation, investigation, and writing- review and editing. L.J., S.L., and T.W. contributed to supervision, project administration, methodology, and writing-review and editing. All authors reviewed, read, and approved the final manuscript.

Funding

This study was supported in part by the Noncommunicable Chronic Diseases-National Science and Technology Major Project under Grant 2024ZD0539500; in part by the National Natural Science Foundation of China under Grants U25A20142 and 62303463; in part by the Beijing Natural Science Foundation under Grants L251071, L246047, L232137, Z241100009024031, and L232137; and in part by the Beijing Nova Program under Grant 20230484336. The funders had no role in the study design, data analysis, or preparation of the manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable. This study was a bibliometric analysis of the published literature and did not involve human participants, human data, or human tissues.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Human and animal rights

Not applicable.