The study analyzed the co-authorship network of researchers in the field of Gastrointestinal Endoscopy, focusing on the periods from 2000 to 2023. The analysis was conducted using data from the WoS and utilized both macro and micro-level network metrics to understand the evolution of collaborative networks in this field.
2000–2009 Network Analysis
During the period from 2000 to 2009, the co-authorship network for Gastrointestinal Endoscopy research comprised numerous components, with a network density of 0.0003654681 (Table 1), indicating a very sparse network (Fig. 1). The average clustering coefficient was 0.908948227 (Table 1), showing a high tendency for authors to form tight-knit groups. The network consisted of 2,194 components (Table 1), indicating a highly fragmented network with many isolated subgroups (Fig. 1). The average distance between nodes was infinite, suggesting that the network was not fully connected, and many authors did not share direct or indirect connections with others 8.
At the micro level, the top 20 authors by degree centrality were identified, with Gasbarrini, G having the highest degree centrality (0.0066), followed by Malfertheiner, P (0.0061) and Di Mario, F (0.0054) (Table 2). These authors were the most connected within the network, collaborating with the largest number of other researchers (Fig. 1). When considering closeness centrality, Vakil, N had the highest score (0.058) (Table 3), indicating that this author was closest to all other nodes in the network, thus occupying a central position in the dissemination of information (Fig. 1). Asaka, M topped the list for betweenness centrality (0.0128) (Table 4), reflecting this Japanese author’s key role in connecting different subgroups within the network (Fig. 1), making them a crucial figure in the flow of information across the network.
2010–2019 Network Analysis
The co-authorship network during the period from 2010 to 2019 displayed a slight decrease in network density, with a value of 0.00026543 (Table 1), indicating an even sparser network compared to the previous decade (Fig. 2). The average clustering coefficient was 0.90408127 (Table 1), similar to the previous period, reflecting a persistent pattern of clustering among authors (Fig. 2). The number of components increased to 3,228 (Table 1), further highlighting the fragmented nature of the network with a significant number of isolated groups (Fig. 2). As in the earlier period, the average distance between nodes remained infinite (Table 1), indicating that many authors were not interconnected within the network.
In terms of micro-level metrics, Hassan, Cesare emerged as the author with the highest degree centrality (0.0112) (Table 2), indicating extensive collaborations with other researchers (Fig. 2). Dumonceau, Jean-Marc (0.0081) and Dinis-Ribeiro, Mario (0.007) followed closely (Table 2), suggesting their influential positions within the network (Fig. 2). Regarding closeness centrality, Dumonceau, Jean-Marc (0.152) and Hassan, Cesare (0.151) were the leading authors (Table 3), occupying central roles within the network, enabling efficient communication and collaboration (Fig. 2). Dumonceau, Jean-Marc also had the highest betweenness centrality (0.0294) (Table 4), indicating a crucial role in bridging different research groups and facilitating the exchange of information across the network (Fig. 2). Fujimoto, Kazuma ranked 12th in degree centrality (0.005). In contrast, Fujishiro, Mitsuhiro Uedo, Noriya, and Gotoda, Takuji ranked 5th and 14th and 18th place in betweenness centrality (0.0174) (Table 4), marking significant contributions from Japanese researchers in this period.
2020–2023 Network Analysis
During the period from 2020 to 2023, the co-authorship network for Gastrointestinal Endoscopy research showed a slight increase in network density, reaching 0.00034 (Table 1), indicating a marginally denser network compared to the previous decade (Fig. 3). The average clustering coefficient rose to 0.913686747 (Table 1), the highest across the three periods, suggesting even stronger clustering tendencies among collaborating researchers (Fig. 3). The number of components decreased to 2,513 (Table 1), indicating some consolidation of the network, though it remained largely fragmented with many isolated subgroups (Fig. 3). The average distance between nodes continued to be infinite, suggesting that the network was not fully connected, and many authors did not share direct or indirect connections with others.
At the micro level, the top 20 authors by degree centrality were identified, with Hassan, Cesare (0.0124), Repici, Alessandro (0.0103), and Dinis-Ribeiro, Mario (0.0098) having the highest degree centrality, showing their prominence in the field (Table 2). These authors were the most connected within the network, collaborating with the largest number of other researchers (Fig. 3). When considering closeness centrality, Hassan, Cesare (0.1607), Dinis-Ribeiro, Mario (0.1596), and Repici, Alessandro (0.1589) (Table 3) had the highest scores, indicating that these authors were closest to all other nodes in the network (Fig. 3), thus occupying central positions in the dissemination of information (Fig. 3). Repici, Alessandro (0.0205), Sharma, Prateek (0.0204), and Hassan, Cesare (0.0164) topped the list for betweenness centrality (Table 4), reflecting their key roles in connecting different subgroups within the network (Fig. 3), making them crucial figures in the flow of information across the network. Japanese researchers maintained a strong presence in this period, with Fujishiro, Mitsuhiro ranking 6th, Tajiri, Hisao 7th, and Saito, Yutaka 9th in betweenness centrality (Table 4), reflecting their ongoing influence in connecting different subgroups within the network. Additionally, Inoue, Haruhiro and Ishikawa, Hideki were ranked 14th and 15th, respectively, underscoring the significance of Japan's contributions to the field.
Summary
The analysis of collaborative networks in Gastrointestinal Endoscopy research from 2000 to 2023 reveals significant insights into the evolution of co-authorship patterns. The study period was divided into three phases: 2000–2009, 2010–2019, and 2020–2023, each exhibiting unique network characteristics.
From 2000 to 2009, the network density was 0.000365468, indicating a sparse network with limited collaborations. The average clustering coefficient was 0.908948227, suggesting a high tendency for authors to form tight-knit groups. The network comprised 2194 components, reflecting a fragmented structure with many isolated groups. The average distance between nodes was infinite, indicating that not all authors were reachable from each other (Table 1). Key authors during this period, based on degree centrality, included Gasbarrini, G., Malfertheiner, P., and Di Mario, F. In terms of closeness centrality, Vakil, N, van Zanten, SV, and Sung, J was prominent. Asaka, M, Adam, V, and Vakil, N were identified as key intermediaries based on betweenness centrality (Table 2–4).
The period from 2010 to 2019 saw a slight decrease in network density to 0.00026543, indicating a further sparsification of the network. The average clustering coefficient remained high at 0.90408127, and the number of components increased to 3228, suggesting an even more fragmented network. The average distance between nodes remained infinite (Table 1). Prominent authors based on degree centrality included Hassan, Cesare, Dumonceau, Jean-Marc, and Dinis-Ribeiro, Mario. Dumonceau, Jean-Marc, Hassan, Cesare, and Sharma, Prateek were notable for their closeness centrality, while Dumonceau, Jean-Marc, Kiesslich, Ralf, and Hassan, Cesare were key intermediaries based on betweenness centrality (Table 2–4).
In the most recent period from 2020 to 2023, the network density was 0.000340538, showing a slight increase compared to the previous decade. The average clustering coefficient was the highest at 0.913686747, indicating a strong tendency for authors to collaborate within clusters. The number of components decreased to 2513, suggesting a more connected network compared to the previous decade. However, the average distance between nodes remained infinite (Table 1). Key authors based on degree centrality included Hassan, Cesare, Repici, Alessandro, and Dinis-Ribeiro, Mario. In terms of closeness centrality, Hassan, Cesare, Dinis-Ribeiro, Mario, and Repici, Alessandro was prominent. Repici, Alessandro, Sharma, Prateek, and Hassan, Cesare were identified as key intermediaries based on betweenness centrality (Table 2–4).
Overall, the analysis highlights the evolving nature of collaborative networks in Gastrointestinal Endoscopy research, with key contributions from both Japanese researchers and internationally recognized figures like Hassan, Cesare (Italy) and Dinis-Ribeiro, Mario (Portugal), who played significant roles across multiple periods. The high clustering coefficients across all periods suggest a strong tendency for authors to form collaborative groups, while the fragmented nature of the network indicates opportunities for further integration and collaboration.