A decade of advancement in electrical stimulation therapy for spinal cord injury: a bibliometric analysis (2015-2025)

Literature sources and search strategy

To ensure a thorough literature review, we first identified key title keywords and augmented them with corresponding MeSH terms from PubMed^[31]. This combined set of terms was used to structure a comprehensive search in the Web of Science Core Collection (Science Citation Index Expanded, SCI-EXPANDED), executed on January 5, 2026. The search string was detailed as follows: TS = (“spinal cord injury” OR “spinal cord injuries” OR “SCI” OR “spinal trauma” OR “spinal lesion” OR “spinal cord damage”) AND (“electrical stimulation” OR “neurostimulation” OR “neural stimulation” OR “functional electrical stimulation” OR “FES” OR “epidural stimulation” OR “transcutaneous electrical stimulation” OR “tES” OR “transcranial electrical stimulation” OR “spinal stimulation”).

The search was performed from January 1, 2015 to December 31, 2025. The explicit inclusion criteria comprised: (1) peer-reviewed original research articles and review papers; (2) studies published within the defined 2015-2025 timeframe; and (3) literature focusing explicitly on electrical stimulation therapies for SCI. The exclusion criteria comprised: (1) meeting abstracts, editorial materials, letters, data papers, news items, conference papers, corrections, and retracted publications; and (2) articles addressing unrelated neurological conditions, non-electrical interventions, or publications not written in English.

All raw search results exported from the database were imported into Microsoft Excel 2019, where a systematic literature deduplication process was independently performed by two researchers to identify and remove identical duplicate records based on exact matches in publication title, author names, publication year, and source journal. Early Access publications were assigned to the year of their initial online publication. Thus, articles first published online in 2025 but subsequently assigned to journal volumes/issues in 2026 were classified as 2025 publications. Following document-type and language screening, deduplication, and title/abstract eligibility assessment, a final set of 1,905 publications was included in the bibliometric analysis. No additional records were excluded during the deduplication or title/abstract screening. The complete retrieval and selection process is depicted in Figure 1.

Figure 1. Schematic diagram of the data retrieval and analysis process.

Data collection and statistics

Annual literature publication numbers and temporal trends

A total of 1,905 articles were screened and selected for analysis. The annual distribution of these publications from 2015 to 2025 was presented in Figure 2. Overall, the annual count generally rose from 2015 to 2025, with a notable recent peak of 216 publications in 2025.

Figure 2. Trend graph of the number of articles published per year in the field.

Analysis of countries/regions

A total of 59 nations worldwide contributed to the 1905 publications on ES for motor function recovery following SCI. The international collaboration network among countries/regions from 2015 to 2025 is visualized in Figure 3 using VOSviewer. Table 1 presents the top 10 countries and institutions in research on ES therapy for SCI. The United States was the most prolific country, accounting for 949 publications (49.82%), followed by China with 264 (13.86%) and Canada with 213 (11.18%). The overwhelming publication share held by the United States is largely driven by robust federal and institutional funding (such as NIH and Department of Defense grants) dedicated to neuroprosthetics, alongside pioneering clinical centers with long-standing trial infrastructures. Meanwhile, China’s rapid rise to second place reflects a substantial recent surge in national investments directed toward neuro-regeneration and advanced rehabilitation engineering. Regarding international cooperative networks, the five countries with the greatest total link strength (TLS) were the United States (TLS = 535), England (TLS = 197), Canada (TLS = 161), Switzerland (TLS = 157), and Italy (TLS = 146). These metrics highlight a dense transatlantic and European collaborative axis, where nations like Switzerland and England act as crucial bridges for multicenter clinical trials and translational bioelectronic studies. The global distribution of research output and international collaboration is presented in Figure 4.

Figure 3. VOSviewer country co-authorship network. Each node represents a country/region; node size reflects publication output, and links indicate collaborative relationships between countries/regions.

Figure 4. Global research output distribution and cooperation.

Analysis of institutions

Among institutions meeting the predefined inclusion threshold, 101 were included in the network analysis. The top ten institutions by publication count were Case Western Reserve University (87 publications), University of Toronto (73 publications), University of Louisville (71 publications), University of Miami (57 publications), Virginia Commonwealth University (56 publications), University of Health Network (53 publications), University of Alberta (52 publications), University of California, Los Angeles (48 publications), University of Sydney (47 publications), and Harvard Medical School (45 publications), as detailed in Table 1. The leadership of institutions such as Case Western Reserve University and the University of Louisville stems from their historical specialization in functional electrical stimulation (FES) and human epidural stimulation trials. These centers possess unique synergies between clinical rehabilitation hospitals and bioengineering laboratories, allowing them to rapidly translate benchtop paradigms into human subjects. The institutional co-authorship network is visualized in Figure 5 using VOSviewer, illustrating the collaborative relationships among the major research institutions in this field. The five institutions with the highest TLS values were the University of Louisville (115), Case Western Reserve University (105), University of California, Los Angeles (102), University of Miami (99), and University of British Columbia (93), demonstrating that high publication volume closely correlates with central positioning in international co-authorship webs.

Figure 5. VOSviewer institution co-authorship network. Each node represents an institution; node size reflects publication output, and links indicate collaborative relationships between institutions.

Analysis of authors and co-cited authors

A total of 111 authors and 90 co-cited authors met the predefined inclusion thresholds and were included in the network analysis. The ten most prolific and frequently cited authors are listed in Table 2, while the author co-authorship network is presented in Figure 6 using VOSviewer, illustrating collaborative relationships among researchers in the field. In terms of publication volume, the top five authors were Gorgey, Ashraf S. (57 publications), Harkema, Susan J. (35 publications), Edgerton, V. Reggie (33 publications), Hamzaid, Nur Azah (26 publications), and Triolo, Ronald J (25 publications), as shown in Table 2. Regarding citation impact, the leading five authors were Gorgey, Ashraf S. (778 citations), Angeli, CA (460 citations), Hofstoetter, US (454 citations), Minassian, K (432 citations), and Harkema, S (344 citations), demonstrating a significant overlap with the most productive researchers. This overlap indicates that primary producers in this field are also the intellectual anchors whose methodologies shape the broader consensus. Analysis of total link strength (TLS) further identified Minassian, K (TLS = 10,456), Hofstoetter, US (TLS = 9,817), Angeli, CA (TLS = 8,995), Sayenko, DG (TLS = 6,853), and Gorgey, AS (TLS = 6,772) as the authors with the strongest overall collaborative links within the co-authorship network (Table 2). Higher TLS values indicate stronger cumulative co-authorship connections with other researchers in the network, suggesting that these authors are highly interconnected within the field. Their strong collaborative connectivity may facilitate the exchange of knowledge, methodologies, and neuromodulation approaches across research groups, thereby contributing to the development and dissemination of electrical stimulation research for SCI.

Figure 6. VOSviewer author co-authorship network. Each node represents an author; node size reflects publication output, and links indicate co-authorship relationships. Shorter distances and denser links indicate closer collaboration.

Analysis of Journals

The top 10 journals and co-cited journals for research on ES therapy for SCI are shown in Table 3. Of the 121 journals included in the analysis, Journal of Spinal Cord Medicine (n = 88, H-index = 84), IEEE Transactions on Neural Systems and Rehabilitation Engineering (n = 50, H-index = 179), and Frontiers in Neuroscience (n = 48, H-index = 186) were identified as the leading three publication venues. Figure 7 presents the VOSviewer network visualization of cited sources (journals), illustrating the citation-related relationships and clustering patterns among influential journals in the field. Together, these journals published 9.76% of the total studies. Among the top ten most productive journals in this domain, half were ranked in the Q1 quartile, representing the top 25% of journals by impact factor. Furthermore, a majority had an impact factor exceeding 3.0, resulting in a collective average impact factor of 3.5. As shown in Table 3, Spinal Cord (H-index = 135) was the most frequently cited journal in this field. According to Journal Citation Reports (JCR) rankings, over half of these top ten journals were also in the Q1 quartile. Seventy percent possessed an impact factor above 3.0, with the group’s average impact factor reaching 7.96.

Figure 7. VOSviewer network visualization of cited sources (journals). Nodes represent journals/sources, and links indicate citation-related relationships among sources; colors denote network clusters.

Analysis of keywords

Keyword occurrence analysis

Table 4 shows the top 20 keywords in research on ES therapy for SCI. A total of 99 author-provided keywords met the predefined inclusion threshold and were included in the keyword co-occurrence analysis, providing an overview of the major research themes and thematic priorities in ES research for SCI. Temporal analysis of author keyword frequency showed a marked evolution in research focus from 2015 to 2025. The ten most frequently utilized keywords are “spinal cord injury” (872), “functional electrical stimulation” (245), “rehabilitation” (193), “electrical stimulation” (188), “neuromodulation” (160), “epidural spinal cord stimulation” (97), “spinal cord stimulation” (81), “neuroprosthesis” (59), “neuromuscular electrical stimulation” (58), and “locomotion” (52). Figure 8 presents the thematic evolution of ES therapy for SCI research topics from 2015 to 2025 based on keyword frequency.

Figure 8. Visual map using VOSviewer network among keywords. Each node represents a keyword, with node size proportional to its occurrence frequency. Node colors and connecting lines represent distinct research clusters and co-occurrence links, mapping the core domains and sub-themes in the field.

Table 4

Top 20 keywords on research of electrical stimulation therapy for spinal cord injury

Rank	Keywords	Counts	Rank	Keywords	Counts
1	spinal cord injury	872	11	neuroplasticity	52
2	functional electrical stimulation	245	12	neurorehabilitation	52
3	rehabilitation	193	13	brain-computer interface	47
4	electrical stimulation	188	14	electromyography	47
5	neuromodulation	160	15	exercise	45
6	epidural spinal cord stimulation	97	16	spasticity	42
7	spinal cord stimulation	81	17	transcutaneous spinal cord stimulation	42
8	neuroprosthesis	59	18	tetraplegia	41
9	neuromuscular electrical stimulation	58	19	stroke	39
10	locomotion	52	20	plasticity	34

Keyword trend topic analysis

To objectively classify the research landscape rather than rely solely on subjective interpretation, the three evolutionary phases were quantitatively delineated using a combination of keyword average publication years (APY), CiteSpace burst detection metrics, and thematic evolution trajectories. Phase 1 (2015-2017) aggregates keywords with earlier APYs and distinct burst profiles centered on foundational safety and secondary complication management. Phase 2 (2018-2021) is quantitatively anchored by peak citation bursts and co-occurrence frequency shifts toward functional rehabilitation and neuroplasticity. Phase 3 (2022-2025) is defined by recent, high-intensity burst metrics and latest-generation keywords associated with precision neuromodulation, brain-spine interfaces (BSI), and artificial intelligence (AI) integration. Figure 9 depicts trend topics based on author keywords. The initial phase of the decade [Phase 1: Foundational Exploration and Symptomatic Management (2015-2017)] focused on the foundational mechanisms of electrical stimulation and the management of secondary complications associated with SCI. Research during this period prioritized clinical safety and basic electrophysiological observations in human subjects, with significant emphasis on addressing autonomic dysfunctions such as neurogenic bladder and promoting secondary healing processes like wound recovery. Furthermore, the emergence of "closed-loop" and "feedback control" indicates that early investigators were already seeking to enhance the precision of stimulation delivery through regulated algorithmic parameters.

Figure 9. Trend topics analysis displaying the temporal distribution and frequency of key terms over the years. Bubble size indicates term frequency, and horizontal bars represent the active time span.

From 2018 to 2021, the field [Phase 2: Functional Restoration and Neural Plasticity (2018-2021)] shifted toward functional recovery and the exploration of neural repair mechanisms. This period marked the peak of research into Functional Electrical Stimulation (FES) and its integration into intensive rehabilitation protocols aimed at restoring complex motor patterns such as gait and locomotion. Crucially, the academic discourse shifted from viewing electrical stimulation as a passive tool for muscle activation to understanding it as a catalyst for neuroplasticity, investigating how targeted electrical intervention can reorganize neural circuits to facilitate sustained functional gains.

The current era [Phase 3: Precision Modalities and Multidisciplinary Integration (2022-2025)] represents a transition toward highly specialized stimulation modalities and the convergence of disparate technologies. Recent trends highlight a move away from generalized electrical therapy toward precise interventions such as Epidural Spinal Stimulation (ESS) and non-invasive Transcutaneous Spinal Cord Stimulation (TSCS), with a growing focus on challenging high-level injuries like cervical SCI. Moreover, the field is increasingly integrating bioelectronics with cutting-edge disciplines, including exoskeleton robotics for enhanced mobility and tissue engineering for biological scaffolding, reflecting a holistic, "biomechanical" approach to spinal cord reconstruction.

Analysis of highly cited references and citation bursts

Based on Web of Science data, 10 references^[37-46] were identified as highly cited and are summarized in Table 5. Separately, Figure 10 presents the top 20 references with the strongest citation bursts detected by CiteSpace. These are distinct bibliometric indicators: Table 5 summarizes citation frequency among selected highly cited references, whereas Figure 10 depicts references showing abrupt increases in citation activity over specific time intervals. The highly cited literature has appeared in leading publications, including The Lancet (IF = 88.5), Brain (IF = 11.7), Nature (IF = 48.5), New England Journal of Medicine (IF = 78.5), and Nature Medicine (IF = 50), underscoring the considerable scientific attention dedicated to ES investigations. Notable highly cited works include Susan Harkema’s paper in The Lancet, Claudia A. Angeli’s article in Brain, and Fabien B. Wagner’s 2018 Nature paper.

Figure 10. Top 20 references with the strongest citation bursts in the field of spinal cord injury and electrical stimulation research from 2015 to 2025. The blue lines indicate the timeline, and the red segments represent intervals during which citation bursts occurred.

The research landscape in electrical stimulation for SCI has evolved from initial proof-of-concept studies to highly sophisticated, multi-modal approaches to functional restoration. Early landmark publications (e.g., Harkema S et al.^[41] in The Lancet and Angeli CA et al.^[38] in Brain) primarily demonstrated the feasibility of restoring volitional movement and autonomic functions in motor-complete patients using repurposed chronic pain devices. In contrast, recent high-impact research (e.g., Wagner FB et al.^[45] in Nature and Angeli CA et al.^[37] in the New England Journal of Medicine) has shifted toward spatiotemporal neuromodulation and BSI. Current hotspots emphasize the restoration of complex overground walking through “targeted” stimulation patterns, the integration of AI for real-time control, and the development of "digital bridges" that synchronize cortical intentions with spinal circuits, marking a transition from general activation to individualized precision medicine.

Overview and hotspots of electrical stimulation therapy in SCI

The bibliometric analysis of 1,905 publications from 2015 to 2025 reveals that ES has transitioned from an experimental adjunct therapy into an increasingly prominent modality in the rehabilitation and neuromodulatory management of SCI. The overall upward trend in annual publications underscores a global consensus that neuromodulation is one of the most promising frontiers for restoring motor function in paralyzed individuals. Geographically and institutionally, the research landscape is anchored by the dominance of the United States (49.82% of publications) and pioneering institutions such as Case Western Reserve University and the University of Louisville, while substantial contributions from China (13.86%) and high total link strength across European nations highlight a collaborative international priority. Furthermore, landmark publications in leading journals such as The Lancet, Nature, and The New England Journal of Medicine highlight the growing scientific attention to ES as a promising neuromodulatory strategy for facilitating functional recovery in individuals with chronic, motor-complete SCI^[37-46].

Clinical significance and translational implications

The trajectory mapped over the past decade suggests an important shift in SCI research: ES is increasingly being investigated not only as an assistive or symptomatic tool but also as a neuromodulatory strategy intended to facilitate neural circuit activation and functional recovery. Traditionally, orthopedic and neurosurgical management of SCI has emphasized structural stabilization, decompression, and rehabilitation. Emerging clinical and experimental studies suggest that residual lumbosacral networks may retain physiological plasticity that can be modulated in selected individuals. Nevertheless, the available evidence remains heterogeneous, and bibliometric trends alone cannot establish disease-modifying effects or clinical efficacy.

By delivering precisely patterned electrical currents, modern neuromodulation effectively elevates the baseline excitability of spinal interneuronal circuits, enabling them to process residual supraspinal signals or interface directly with cortical intents via digital bridges. This transforms the clinical objective from passive compensation to active functional recovery. For the broader orthopedic and neurosurgical communities, these findings signal a vital change in clinical mindset. The future of SCI rehabilitation will not rely on finding a single biological or mechanical cure, but rather on fostering a multidisciplinary ecosystem where bioelectronics, robotics, and tissue engineering converge. In this integrated framework, electrical stimulation serves as the primary catalytic driver, bridging the gap between surgical reconstruction and functional independence^[47-51].

Future outlook

Over the past decade, the technological landscape of ES for SCI has undergone a fundamental transformation. Emerging technologies have moved beyond simple muscle activation toward high-precision, cell-specific modulation and "smart" closed-loop systems that integrate advanced materials with biological repair strategies.

Limitations of this study

Several limitations must be acknowledged. First, regarding database selection and retrieval strategy biases, our data source was restricted to the Web of Science Core Collection to maintain consistency in citation metrics and metadata formatting; this inherently omits relevant literature indexed in alternative databases such as Scopus, Embase, or PubMed. Additionally, our specific search strings and Boolean operators, while comprehensive, may have introduced retrieval strategy bias by missing eligible publications that utilized variant terminology, alongside language restrictions (focusing solely on English-language articles and reviews) and the exclusion of grey literature or conference abstracts, which may narrow our global perspective on emerging research trends. Second, concerning methodological constraints in citation analysis, citation counts can be influenced by citation bias, self-citations, and the passage of time (favoring older publications regardless of contemporary relevance), while bibliometric tools primarily evaluate quantitative publication trends, citation networks, and structural mappings rather than the underlying methodological rigor or true scientific quality of individual studies. Finally, regarding clinical translation and technical heterogeneity, bibliometric evaluations cannot directly assess the clinical efficacy, safety profiles, or therapeutic superiority of specific interventions. The dataset aggregates a broad spectrum of distinct neuromodulation modalities - such as epidural electrical stimulation, transcutaneous spinal cord stimulation, and functional electrical stimulation—that exhibit significant technical heterogeneity, varying mechanisms of action, and diverse clinical indications, which cannot be fully differentiated through macro-level bibliometric mapping alone.

Conclusions

This bibliometric analysis of the literature from 2015 to 2025 underscores a transformative decade in ES for SCI. The evidence demonstrates that ES has advanced from basic muscle activation toward more sophisticated spatiotemporal and closed-loop neuromodulation paradigms.

The primary findings of this study highlight three critical shifts in the research landscape. First, technology has matured significantly, moving from generalized FES to high-precision modalities such as ESS and TSCS. Second, the integration of AI and BSI has effectively created a “digital bridge”, restoring volitional control by synchronizing cortical intentions with spinal circuit modulation. Finally, the therapeutic scope has expanded beyond motor recovery to include the vital restoration of autonomic functions, such as hemodynamic stability and neurogenic bladder management, which are essential for the holistic recovery of SCI patients.

While these findings reflect a decade of profound progress, the scope of this bibliometric analysis is subject to certain limitations, including potential database bias and the exclusion of non-English literature, which may restrict the universality of the global trends identified. Furthermore, challenges remain in standardizing stimulation protocols and improving access to invasive technologies. Nonetheless, given current methodological limitations, including study heterogeneity and small sample sizes, future research must prioritize rigorous, standardized clinical trials to fully establish its long-term therapeutic efficacy and translational potential.

In conclusion, the past ten years have laid the foundation for a new era of personalized neuromodulation. For clinicians and researchers alike, the focus must now shift toward accelerating clinical translation, ensuring that these high-impact scientific breakthroughs reach the bedside to fundamentally improve the lives of those affected by SCI.