2.1 Data source and retrieval strategy

To ensure data authority and comprehensiveness, the Science Citation Index Expanded (SCI-Expanded) within the Web of Science Core Collection (WoSCC) was selected as the core platform for literature retrieval. This database is globally recognized for its rigorous inclusion criteria and broad disciplinary coverage, establishing it as a high-quality source of academic literature.^15–17 For retrieval strategies, the study adhered to the Medical Subject Headings (MeSH) terminology system, combining Entry Terms to construct a scientific and comprehensive search expression: (‘knee osteoarthritis’ OR ‘osteoarthritis of knee’ OR ‘osteoarthritis of the knee’) AND (biomechanic∗ OR mechanobiolog∗ OR kinematics OR ‘finite element’). This approach balances retrieval precision with broad coverage by integrating core keywords and extended vocabulary. To enhance data relevance, the search results were subjected to dual restrictions: a time frame limited to 2004–2023 to reflect recent academic trends and inclusion restricted to Research Articles and Review Articles, excluding non-core types such as conference papers and editorial comments. Additionally, all analyzed literature was restricted to English to ensure data consistency and comparability. To mitigate inconsistencies from database updates, literature retrieval and data extraction were completed on the same day. This strategy ensured data timeliness and minimized potential sample bias. After rigorous screening, irrelevant literature was excluded, resulting in a final sample of 1932 high-quality articles. This sample size ensures statistical power while maintaining data refinement, providing a robust foundation for subsequent knowledge mapping and bibliometric analysis.

2.2 Bibliometric analysis

Bibliographic data were exported from the database and saved in two formats: plain text files (named “download_xxx.txt") and tab-delimited files. Data deduplication was performed using CiteSpace 6.2.R4, confirming the absence of duplicate records. Following preprocessing, scientific knowledge mapping and statistical analysis were conducted. Multiple analytical tools were integrated: VOSviewer 1.6.19 for visualization, CiteSpace 6.2.R4 for temporal and node importance analysis, and Pajek 5.18 for complex network processing. In VOSviewer, thresholds were meticulously calibrated: minimum publication counts of 5 for countries/regions, 20 for institutions, and 10 for authors; minimum citation thresholds of 300 for journals, 50 for documents, and 100 for authors; keywords required a minimum of 50 occurrences. Tab-delimited files were uploaded to an online bibliometric platform to construct knowledge maps of international collaboration networks, aiming to elucidate global patterns of academic collaboration and knowledge flow. In CiteSpace, the analysis spanned January 2004 to December 2023, with 2-year time slices. Node types included keywords, categories, and references, with the top 50 nodes per slice selected based on occurrence or citation frequency. Default parameters were retained to ensure analytical consistency and accuracy.

3.1 Analysis of annual publications and citations

Research output and citation frequency in the field have increased exponentially over the past two decades. As depicted in Fig. 1, annual publications grew from 26.4 articles (2004–2008) to 165 articles (2019–2023), marking a 6.25-fold increase, while citation frequency rose from 148 to 6103.8, an increase of 41.24-fold. This surge highlights the ongoing emergence of research hotspots and sustained academic engagement. The cumulative citation frequency reached 55,739, with an average of 28.85 citations per document and an H-index of 96, underscoring the field's significant academic impact. A total of 1932 articles were systematically analyzed, with empirical studies accounting for 87.53 % (1691 articles) and review articles representing 12.47 % (241 articles). These publications collectively form the theoretical foundation of the field and have achieved broad academic recognition. The predominance of empirical research reflects robust data support for theoretical advancement, while review articles provide a macroscopic perspective, enhancing knowledge integration and dissemination. Despite rapid growth in volume and impact, the field confronts dual challenges: theoretical depth and practical complexity. Globally collaborative innovation has accelerated theoretical development, yet breakthrough progress demands interdisciplinary methodologies and experimental paradigm shifts. This dynamic balance indicates the field is at a critical juncture of transitioning from quantitative expansion to qualitative advancement. Future research should prioritize the integration of theory and practice to address complex challenges and foster sustained disciplinary evolution.

Close

Fig. 1 Trends in publications and citations.

Export to PPT

3.2 Analysis of countries/regions and institutions

This study employs a global bibliometric analysis, integrating academic outputs from 1986 research institutions across 69 countries/regions. Fig. 2 visualizes the dynamic trajectories of annual publication outputs from the top ten producing countries. The data reveal that the United States has maintained a dominant position in the field over the past two decades. In contrast, China, following an initial period of gradual growth, experienced leapfrog growth starting in 2020, driven by national strategic initiatives and increased research investment, and now consistently ranks as the second most productive country globally. Fig. 3 presents an international collaboration network, where node size and link density intuitively reflect research volume and cooperation patterns.¹⁸ The United States not only dominates in output scale but also demonstrates exceptional international collaboration capacity, with its network spanning major research hubs across five continents. Multidimensional visualization in Fig. 4, utilizing a node-link-color coding system, further illustrates the temporal depth and spatial connectivity of research capabilities.^19,20 This analysis highlights the foundational role of traditional research powers (the U.S., Canada, Australia, the U.K., and Germany) alongside China's rapid emergence as a rising force. Quantitative metrics in Table 1 corroborate this landscape: the United States leads with 700 publications (36.23 % of the total), contributing over half of the research outputs alongside Canada (269 publications, 13.92 %). The U.S. also excels in core indicators such as total citations (TC), H-index, and total link strength (TLS), underscoring its global radiative impact and deeply embedded collaborative networks. Notably, the U.S. forms a “knowledge-sharing hub” with Canada, Australia, the U.K., Germany, and China, serving as a critical driver of disciplinary advancement. While the Netherlands ranks eighth in publication volume, its average citation per paper (ACPP) is the highest, reflecting high-impact research. China, though prominent in quantitative metrics, shows room for improvement in quality-oriented indicators such as ACPP. These findings suggest that China's research ecosystem requires a transition toward a balance of quantity and quality. Strategies such as deepening international collaborations, adopting cutting-edge methodologies, and optimizing resource allocation could enhance innovation and global radiative impact. Future efforts should prioritize the development of quality-oriented evaluation systems to shift from high-volume production to high-impact contributions, thereby securing a more dominant academic position in the global research ecosystem. Such a transformation not only affects individual nations' academic reputations but also shapes the geographic landscape of global knowledge production and disciplinary paradigms.

Close

Fig. 2 Trends in publications from the top 10 high-yield countries/regions.

Export to PPT

Close

Fig. 3 Knowledge map of national/regional cooperation networks.

Export to PPT

Close

Fig. 4 Time overlay of national/regional cooperation networks.

Export to PPT

In the analysis presented in Fig. 5, the structural characteristics of collaboration networks among institutions with 20 or more publications are examined, alongside systematic insights into the distribution patterns of their average publication timelines. Table 2 provides detailed data on key academic metrics for the top ten most productive institutions. During the early stages of the discipline's development, institutions such as the University of Melbourne, the University of California System, Harvard University, Boston University, and Stanford University established foundational knowledge frameworks through their early academic outputs. In recent years, institutions like the University of North Carolina, KU Leuven, and the University of Eastern Finland have emerged as rising forces, demonstrating notable contributions to research productivity and impact. Specifically, the University of Melbourne leads with 93 publications, accounting for 4.81 % of the total output, reflecting significant academic leadership in the field. The University of California System and Harvard University follow closely, with 61 (3.16 %) and 60 (3.11 %) publications, respectively, reinforcing their core positions at the disciplinary frontier. Although Stanford University ranks sixth in total publications, it tops both TC and ACPP, underscoring the high quality and broad recognition of its research. This further solidifies its status as a central research institution in the field. Meanwhile, the Veterans Health Administration demonstrates exceptional performance in TLS, highlighting its extensive international collaborative networks. Notably, most high-productivity institutions are concentrated in the United States, reaffirming the country's leadership and substantial contributions to global academic development. However, Chinese institutions have yet to feature in the top ten for high productivity and impact, indicating the need for continued efforts to enhance the quality of academic outputs, strengthen international collaborations, and expand global influence. This underscores the necessity for Chinese research institutions to intensify international partnerships, optimize resource allocation, and boost research innovation to elevate their competitiveness and influence in the global academic ecosystem.

Close

Fig. 5 Time overlay of institutional collaboration networks.

Export to PPT

3.3 Analysis of authors

In the academic ecosystem of this research field, a total of 6929 researchers were documented as contributing to scholarly output. Fig. 6A visualizes the collaborative network structure and time-series distribution of scholars who contributed at least ten publications, while Table 3 systematically compiles bibliometric indicators for the top ten most active authors in the field. Bibliometric analysis revealed that Professor Bennell KL from the University of Melbourne, Australia, ranked first with 47 publications (2.43 % of the total output), demonstrating exceptional academic productivity. Hinman RS, also from the University of Melbourne, ranked second with 40 publications (2.07 %), followed by Hunt MA from the University of British Columbia, Canada, and Wrigley TV from the University of Melbourne, both with 35 publications (1.81 % each). These findings highlight the robust academic clusters in the Pacific Rim region. Notably, Bennell KL not only led in quantity but also topped the TLS index, reflecting his exceptional ability to establish cross-institutional collaborations. His stable partnerships with core team members such as Hinman RS, Wrigley TV, Metcalf BR, and Hall M emerged as critical nodes in knowledge production within the field. Although Andriacchi TP from Stanford University ranked fifth with 30 publications, his performance in quality metrics such as TC, ACPP, and H-index was among the highest, underscoring the high impact and translational efficiency of his research. Historical analysis indicated that senior scholars like Bennell KL and Hinman RS played foundational roles in establishing theoretical frameworks during the formative stages of the discipline, with their scholarly legacy continuing to exert a long-tail effect. In contrast, emerging scholars such as Pietrosimone B and Charlton JM contributed innovative vitality through frontier explorations, creating a complementary intergenerational knowledge production dynamic. Geographical distribution analysis revealed that high-impact authors were predominantly concentrated in top research universities in Australia and the United States, reflecting a pronounced polarization in knowledge production. To address the absence of Chinese scholars in the global high-productivity cohort, this study proposes the establishment of a “pyramid-style” talent cultivation framework: foundational research capacity building at the base, international collaboration network strengthening in the middle tier, and academic leadership development at the top. By implementing targeted academic support policies, refining research evaluation mechanisms, and encouraging original research, it is anticipated that a domestically rooted academic community with international influence could emerge within the next 5–10 years, transitioning from knowledge consumers to knowledge producers.

Close

Fig. 6 (A) Time overlay of author's cooperative networks. (B) Author co-citation network knowledge map.

Export to PPT

In an in-depth analysis of academic knowledge dissemination networks, Fig. 6B employs visualization techniques to reveal the complex co-citation ecology among scholars with citation frequencies ≥100. The figure adopts a multidimensional encoding strategy: node size correlates positively with citation frequency, lines between nodes indicate co-citation relationships, and color coding reflects academic community divisions based on thematic similarity.^21,22Table 4 presents bibliometric data showing that Andriacchi TP ranked at the core hub position with 787 citations, followed by Felson DT (718) and Sharma L (590). These figures not only quantify academic influence but also reflect the structural role of their research within the knowledge production system. Further network analysis revealed that Andriacchi TP demonstrated exceptional performance in the TLS index, indicating his irreplaceable role in building interdisciplinary knowledge bridges.²³ His highly connected citation network, formed with key nodes such as Felson DT, Sharma L, Mündermann A, Bennell KL, and Kellgren JH, constitutes the primary channel for knowledge diffusion in the field, illustrating the collaborative effects of the academic community under specific research paradigms. Notably, despite China's significant share in global research output, it has yet to establish representative academic nodes within this high-frequency citation network. This phenomenon suggests that current academic evaluation systems should shift from a quantity-driven paradigm to one prioritizing quality. Strategic pathways to enhance the international academic voice of Chinese scholars include: Establishing a dual-driven evaluation system based on quality-impact incorporating innovation, methodological rigor, and international citation potential as core metrics. Promoting deep international collaborations through joint research laboratories and transnational academic alliances to foster knowledge co-creation. Optimizing dissemination mechanisms by encouraging paradigm-shifting research in top-tier international journals. Cultivating academic leaders through systematic academic diplomacy strategies to increase the representation of domestic scholars in international academic organizations. Implementing these strategies could cultivate a Chinese academic community with both local characteristics and global influence within the next academic cycle, thereby constructing China-specific academic growth poles in the international knowledge production system and promoting the diversification and coexistence of disciplinary ecosystems.

3.4 Analysis of journals

This study systematically analyzed 355 academic journals to uncover the core publication landscape and academic geography of the field. The findings revealed a pronounced geographic concentration of global academic resources, with European and North American journals dominating, particularly those from the UK, which accounted for 50 % of the total. These journals serve as critical hubs for advancing the discipline and facilitating international academic exchange. This geographic distribution is not only reflected in publication volume but also permeates multidimensional indicators of academic influence, underscoring the central role of Europe and North America in knowledge dissemination and academic innovation. In terms of publication volume, Gait Posture ranked first with 148 articles (7.66 % of the total), establishing itself as the flagship journal of the discipline. Following closely were Clin Biomech (132 articles, 6.83 %) and J Orthop Res (112 articles, 5.80 %), forming the second tier of high-output journals. Quantitative data in Table 5 showed that these high-output journals had an average H-index of 157.75 and an average impact factor (IF) of 3.13, with the majority ranked in Journal Citation Reports (JCR) Q1/Q2, confirming their high academic value and global reach. However, publication volume is not the sole measure of academic influence. While Osteoarthr Cartilage ranked fourth in publication volume, it topped multidimensional evaluation metrics, including TC, ACPP, and IF. This highlights its deep academic legacy and broad disciplinary influence.

Fig. 7 illustrates a complex co-citation network, focusing on journals with citation frequencies of 300 or higher, revealing intricate academic connections among them. Table 6 presents core data for the ten most highly cited journals in the field, underscoring their prominent academic standing. Among these, Osteoarthr Cartilage boasted the highest citation frequency of 5,496, establishing its dominance in the field. Following closely were J Biomech and Gait Posture, with citation frequencies of 4859 and 3,144, respectively, reflecting their substantial academic contributions and broad recognition. Notably, Osteoarthr Cartilage not only led in citation frequency but also achieved the highest TLS value, indicating its central role in academic discourse through extensive interactions with other high-impact journals such as Arthritis Rheum-US, Ann Rheum Dis, J Biomech, Am J Sport Med, J Orthop Res, J Bone Joint Surg Am, Clin Biomech, Gait Posture, and Clin Orthop Relat R. While Ann Rheum Dis ranked seventh in citation frequency, its IF placed it at the top, reaffirming its status as a premier academic journal. A striking observation is that all ten journals originated from western countries, with five based in the United Kingdom. This geographic concentration highlights the pivotal role of Western nations, particularly the UK, in advancing knowledge dissemination and academic innovation within the discipline. The analysis underscores the interconnected nature of these journals, emphasizing their collective contribution to shaping the field's intellectual landscape.

Close

Fig. 7 Journal co-citation network knowledge map.

Export to PPT

Fig. 8 presents a citation interaction network, with citing source journals on the left and cited target clusters on the right.^24,25 This visualization highlights the interdisciplinary nature of knowledge flow, with six primary citation pathways illustrated through color-coded connections.^26,27 These pathways reveal how journals in fields such as molecular/biology/genetics, health/nursing/medicine, and sports/rehabilitation/sport serve as critical knowledge sources for journals in domains like medicine/medical/clinical and neurology/sports/ophthalmology. The analysis offers two key insights for academic strategy development. First, high-impact journals are identified as critical sources for tracking disciplinary frontiers. Second, submission decisions should prioritize active contributor journals to accelerate knowledge dissemination and enhance academic influence. Together, these strategies aim to foster a more efficient global academic exchange ecosystem. This structured approach underscores the importance of interdisciplinary citation networks in advancing scholarly discourse and optimizing research impact.

Close

Fig. 8 Journal dual-map overlay.

Export to PPT

3.5 Analysis of subject categories

This study systematically integrated information from 93 academic disciplines through an interdisciplinary lens to construct a knowledge network for the field of KOA biomechanics. Fig. 9 visually illustrates the structural characteristics of the disciplinary interaction network, with a filtering criterion of a minimum frequency of 10 occurrences per discipline. The size of the nodes reflects the frequency of each discipline, the lines represent the strength of connections between disciplines, and the depth of color reveals the dynamic evolution of disciplines over time.^28,29 This multidimensional visualization approach enhances data readability and provides a basis for identifying synergistic effects between disciplines. By incorporating betweenness centrality (BC) metrics (with a threshold of 0.1), the study identified key hub nodes in the network, marked by purple halos, which play a critical role in network connectivity and serve as central nodes for knowledge flow.^30,31 For instance, Orthopedics ranked first with a BC of 0.66, underscoring its irreplaceable core position in multidisciplinary integration. Drawing on data from Fig. 9 and Table 7, the study quantified the synergistic effects between disciplines. Orthopedics dominated the research landscape with 963 occurrences, accounting for 49.84 % of the total frequency, establishing itself as the most dynamic research hotspot. This high frequency not only highlights its central role in KOA research but also reflects its sustained contribution to knowledge production. Sport Sciences and Engineering Biomedical followed with 678 occurrences (35.09 %) and 394 occurrences (20.39 %), respectively, forming the foundational pillars of the research domain. This distribution indicates that KOA research has evolved into a stable structure centered on Orthopedics, supported by multidisciplinary collaboration. Mathematical Computational Biology and Surgery ranked second and third in BC, with scores of 0.51 and 0.47, respectively, demonstrating their significant role in constructing interdisciplinary research networks. By bridging knowledge systems across disciplines, these fields have facilitated deeper integration of research efforts.

Close

Fig. 9 Category co-occurrence network knowledge map.

Export to PPT

Based on the aforementioned analysis, future research should prioritize Orthopedics as the central discipline, further amplifying its leadership in elucidating disease mechanisms, refining diagnostic criteria, and advancing therapeutic strategies. For instance, integrating biomechanical principles with clinical datasets could facilitate the development of early-warning models grounded in mechanical parameters. Additionally, enhanced research investment in bridging disciplines such as Mathematical Computational Biology and Surgery is critical to reinforcing their role as hubs within interdisciplinary networks. Computational biology models, for example, could be employed to refine surgical protocols and improve clinical outcomes. Furthermore, deeper engagement of Sport Sciences and Engineering Biomedical disciplines should be encouraged, particularly in applications related to rehabilitative biomechanics and modeling. Studies in sports biomechanics, for instance, could optimize rehabilitation training protocols and enhance functional recovery efficiency in patients. To facilitate these advancements, dynamic interdisciplinary collaboration mechanisms should be established. Strategies may include organizing regular academic forums, developing shared data platforms, and instituting interdisciplinary research funds to support multi-disciplinary teams in joint projects. This strategic optimization aligns with the growing emphasis on interdisciplinary research in academia and provides a clear roadmap for sustainable research development.

3.6 Analysis of highly cited references

In the field of KOA biomechanics, highly cited literature has established the theoretical foundation for disciplinary development.^32,33 By analyzing these seminal works, the intrinsic logic of disciplinary evolution becomes clear: from the establishment of diagnostic criteria to the elucidation of biomechanical mechanisms, and ultimately to risk prediction and intervention strategies (Fig. 10). The following analysis framework, restructured by research themes, reveals how these milestone studies progressively advanced academic progress (Table 8). In 1957, Kellgren JH and colleagues³⁴ published a groundbreaking study in Ann Rheum Dis, quantifying observer agreement in radiographic interpretation of osteoarthritis. Significant consistency was demonstrated across all joints except the wrist. This finding directly led to the 1986 development of osteoarthritis classification criteria by Altman R and colleagues³⁵ in Arthritis Rheum-US, which revolutionized diagnosis by replacing subjective clinical experience with standardized criteria. Together, these two studies laid the foundation for modern KOA diagnostics, remaining indispensable cornerstones in the field. In 1998, Sharma L and colleagues,³⁶ in Arthritis Rheum-US, established a significant association between knee adduction moment (KAM) and medial tibiofemoral osteoarthritis severity, linking biomechanical parameters to quantitative disease indicators for the first time. Earlier, in 1991, Schipplein OD(³⁷), in J Orthop Res, analyzed compensatory gait mechanisms in patients with knee varus deformity, revealing increased flexion-extension torque to maintain balance. These studies introduced biomechanical analysis into KOA pathomechanisms, establishing a novel research paradigm. In 2002, Miyazaki T and colleagues,³⁸ in Ann Rheum Dis, demonstrated the predictive value of baseline KAM for radiographic progression of medial compartment KOA, pioneering biomechanics-based disease prediction. That same year, Baliunas AJ(³⁹), in Osteoarthr Cartilage, reported elevated lateral KAM in KOA patients, further underscoring the clinical relevance of torque parameters. These studies established dynamic monitoring as central to disease management. In 2005, Mündermann A and colleagues,⁴⁰ in Arthritis Rheum-US, revealed a positive correlation between KOA severity and peak KAM, alongside reduced hip adduction moment in matched-control studies. This cross-joint biomechanical association provided empirical support for multidimensional severity assessment. Subsequently, in 2008, Astephen JL and colleagues,⁴¹ in J Orthop Res, demonstrated dose-response relationships between KOA severity and multisegmental joint dynamics through gait analysis, refining disease quantification. In 2004, Andriacchi TP and colleagues,⁴² in Ann Biomed Eng, integrated biomarkers, cartilage morphology, and gait analysis into a pathomechanical framework, offering a systematic approach to disease intervention. Finally, in 2011, Bennell KL and colleagues,⁴³ in Ann Rheum Dis, confirmed a direct relationship between KAM impulse and medial tibial cartilage volume loss in longitudinal studies, positioning biomechanical parameters as key targets for disease-modifying therapies. In summary, these highly cited studies form an interconnected chain—from standardization to mechanistic insights, predictive modeling, and intervention strategies—constructing a systematic knowledge framework for KOA research. Their academic authority, reflected in Q1/Q2 journal placements and cross-temporal citation networks, maps the intellectual trajectory of the field. These works are not merely symbols of academic achievement but enduring catalysts for disciplinary advancement, providing a robust theoretical foundation and methodological guidance for future research.

Close

Fig. 10 Reference co-citation network knowledge map.

Export to PPT

3.7 Analysis of references burst

Citation burst phenomenon, a critical concept in scientific research, serves as a robust indicator of emerging trends by reflecting substantial increases in citation frequency within specific timeframes.^44,45 Through application of a 3-year minimum burst duration threshold, we identified the top thirty publications with the most pronounced citation bursts (Fig. 11). The “Strength" metric quantifies burst intensity, with higher values indicating greater significance.⁴⁶ Temporal parameters ("Begin" and “End") demarcate burst initiation and termination, visually represented by blue bars for total duration and red segments highlighting active burst periods.^47–49 Notably, twelve publications maintained citation bursts through 2023 or beyond, establishing them as pivotal markers of current research priorities and future directions. Key mechanistic insights were revealed through these studies: Investigations by Chehab EF et al.⁵⁰ (2014) demonstrated differential impacts of KAM and knee flexion moment (KFM) on femoral versus tibial cartilage degeneration, providing novel perspectives for osteoarthritis intervention strategies and risk assessment protocols. Chang AH et al.⁵¹ (2015) established significant correlations between baseline peak KAM/KAM impulse magnitude and progression of medial bone marrow lesions coupled with cartilage thickness reduction, thereby identifying critical biomechanical targets for decelerating medial compartment degeneration. Notably, longitudinal analysis by Wellsandt E et al.⁵² (2016) revealed that reduced KAM and medial compartment contact forces during early post-anterior cruciate ligament reconstruction gait were associated with radiographic KOA development at 5-year follow-up, suggesting potential early-warning biomechanical markers. The psychometric validation conducted by Roos EM et al.⁵³ (1998) confirmed the Knee injury and Osteoarthritis Outcome Score (KOOS) as a reliable, cost-effective patient-reported outcome measure for both short- and long-term evaluation of post-traumatic knee pathology. Biomechanical correlations were further elucidated by Kutzner I et al.⁵⁴ (2013), who identified strong lateral KAM-medial tibiofemoral contact force associations during early stance phase but diminished correlations in late stance, highlighting the need for multi-factorial predictive models. Felson DT's⁵⁵ (2013) comprehensive review redefined osteoarthritis pathogenesis as mechanically-induced joint failure, advocating for biomechanical correction over anti-inflammatory approaches in therapeutic decision-making. Clinical practice guidelines received substantial updates through Bannuru RR et al.'s⁵⁶ (2019) Osteoarthritis Research Society International (OARSI) recommendations, which introduced patient-centered treatment algorithms for multi-joint osteoarthritis management. The epidemiological analysis by Cross M et al.⁵⁷ (2014) emphasized the underestimated global burden of hip/knee osteoarthritis, projecting escalating healthcare demands due to demographic shifts and obesity pandemics. Mechanistic investigations by Heiden TL et al.⁵⁸ (2009) identified characteristic kinematic/kinetic alterations in KOA gait patterns that correlated with pain symptoms, suggesting potential neuromuscular compensation strategies through lateral muscle activation. Counterintuitively, Mills K et al.⁵⁹ (2013) reported no consistent KAM variations across disease severity strata, underscoring the necessity for refined biomechanical classification systems. Pathophysiological complexity was addressed in Loeser RF et al.'s⁶⁰ (2012) multi-tissue paradigm, integrating mechanical, inflammatory, and genetic factors in osteoarthritis progression while advocating for multi-omics approaches in disease-modifying therapy development. Clinical applicability was enhanced through Cheung RTH et al.'s⁶¹ (2018) demonstration of gait retraining efficacy for immediate KAM reduction and symptomatic improvement in early KOA patients. Collectively, these seminal works provide multidimensional insights into osteoarthritis mechanisms spanning biomechanical determinants, assessment methodologies, therapeutic interventions, and population health implications. Future research directions should emphasize interdisciplinary integration of mechanical and biological pathways, coupled with development of targeted interventions to modify disease trajectories. Particular emphasis should be placed on longitudinal biomechanical monitoring, personalized treatment algorithms, and cost-effective preventive strategies to address this escalating global health challenge.

Close

Fig. 11 References burst map.

Export to PPT

3.8 Analysis of keywords

The systematic analysis of high-frequency keywords serves as a critical methodological framework for distilling essential research themes and tracking disciplinary evolution within specialized academic domains.^62,63 As shown in Table 9, the top 20 high-frequency keywords reveal the important research content in this field. By integrating co-occurrence network mapping, cluster analytics, and burst strength evaluation, a comprehensive understanding of research priorities, emerging frontiers, and developmental trajectories can be achieved. First, keyword co-occurrence networks provide spatial visualization of conceptual breadth and thematic interconnectedness. As illustrated in Fig. 12, lexical items with occurrence frequencies ≥50 were mapped, revealing pronounced temporal relevance through red-hued nodal prominence (e.g., total knee arthroplasty (TKA), outcomes, injury, baseline, exercise, and meniscus). These visually salient terms reflect current investigative priorities and temporal shifts in research focus. Second, cluster analysis of keywords elucidates the dynamic evolution of research themes. Fig. 13 presents an log-likelihood ratio (LLR) algorithm-derived taxonomy comprising 18 statistically significant clusters, temporally contextualized through timeline visualization. The exceptional clustering validity metrics (modularity Q = 0.8546; weighted mean silhouette S = 0.937) confirm robust classification integrity.^64,65 The persistent dominance of Cluster #1 (disease progression) across temporal strata underscores sustained scholarly emphasis on longitudinal pathological mechanisms. Finally, keyword burst analysis enables precise identification of transient research hotspots and paradigm shifts.^66–68 Through implementation of a 3-year burst duration threshold, thirty high-intensity burst terms were identified (Fig. 14). These temporally constrained bursts not only demarcate period-specific investigative priorities but also serve as predictive indicators when extending beyond contemporary timelines. Notably, terms including outcomes, total knee arthroplasty, people, MRI, and risk exhibited burst persistence through 2023 or later, establishing them as dual markers of current relevance and future directional signposts for osteoarthritis research.

Close

Fig. 12 Keyword co-occurrence network time overlay map.

Export to PPT

Close

Fig. 13 Keyword timeline map.

Export to PPT

Close

Fig. 14 Keywords burst map.

Export to PPT

Contemporary advancements in TKA research demonstrate substantial progress in quality-of-life enhancement through methodological refinements. Prosthetic design and surgical techniques have been systematically optimized via biomechanical analyses, resulting in improved patient satisfaction and functional outcomes. These innovations are particularly evidenced by reduced postoperative complications and enhanced joint kinematics in weight-bearing activities. The evaluation of therapeutic outcomes and injury mechanisms has been rigorously pursued through comparative effectiveness studies. Multidimensional assessments of conservative versus surgical interventions have elucidated biomechanical determinants of treatment efficacy, particularly regarding gait restoration and pain modulation. Concurrently, mechanistic investigations into traumatic and degenerative joint injury pathways have informed evidence-based clinical decision matrices. Baseline parameter establishment and exercise intervention protocols have emerged as critical components in early-stage KOA management. Standardized diagnostic frameworks incorporating quantitative mobility metrics are being developed, while structured rehabilitation programs demonstrating efficacy in delaying disease progression have been validated through randomized controlled trials. These approaches synergistically enable risk-stratified preventive strategies and personalized therapeutic regimens. Cutting-edge investigations into meniscus biomechanics, MRI-based diagnostic algorithms, and risk factor stratification are reshaping KOA pathogenesis understanding. High-resolution imaging modalities now permit non-invasive quantification of meniscal load distribution patterns, while machine learning-enhanced MRI analytics enable early detection of subchondral bone alterations. Epidemiologic meta-analyses have further identified modifiable risk clusters ranging from metabolic comorbidities to occupational biomechanical stressors. Collectively, this keyword-driven analytical paradigm not only synthesizes historical knowledge but also establishes predictive frameworks for future inquiry. The integration of computational modeling, longitudinal cohort studies, and precision medicine approaches is anticipated to drive transformative innovations in musculoskeletal research. Particular emphasis should be placed on translating biomechanical insights into cost-effective clinical protocols while addressing health disparities in arthroplasty accessibility.

3.9 Strengths and limitations

A bibliometric methodology integrated with visual knowledge mapping techniques was systematically implemented to analyze the biomechanical research landscape of KOA. This approach demonstrates substantial advancements in both data comprehensiveness and analytical precision compared to conventional narrative reviews or meta-analyses.⁶⁹ Data acquisition was restricted to the SCI-Expanded subset within WoSCC, ensuring the authority and standardization of source materials. While this selection protocol may exclude non-core journal publications, it effectively mitigates contamination from low-quality data, thereby establishing a robust analytical foundation. Objective quantification of citation networks, co-occurrence patterns, and knowledge structures was performed to eliminate subjective biases inherent in traditional literature synthesis. The methodological framework not only enhances result reliability but also provides a systematic architecture for constructing domain-specific knowledge graphs in KOA biomechanics. While methodological rigor was maintained, several limitations warrant acknowledgment. First, the inclusion criteria (e.g., document type and language restrictions), though quality-enhancing, may induce unintended information loss through exclusion of eligible studies. Second, academic impact assessment based on time-accumulated citation metrics inherently disadvantages recent high-quality publications, potentially obscuring cutting-edge advancements. Despite constraints imposed by data source limitations, temporal citation dynamics, and recency biases, the methodological validity and result robustness retain significant scholarly value. Future investigations should incorporate expanded data repositories (e.g., Scopus, Embase) and machine learning algorithms to dynamically monitor emerging research frontiers. Such methodological syntheses are anticipated to provide sustained theoretical frameworks and practical guidance for KOA biomechanical research, particularly through integration of real-time analytics and predictive modeling capabilities.