Dataset Information

Can co-authorship networks be used to predict author research impact? A machine-learning based analysis within the field of degenerative cervical myelopathy research.

ABSTRACT:

Introduction

Degenerative Cervical Myelopathy (DCM) is a common and disabling condition, with a relatively modest research capacity. In order to accelerate knowledge discovery, the AO Spine RECODE-DCM project has recently established the top priorities for DCM research. Uptake of these priorities within the research community will require their effective dissemination, which can be supported by identifying key opinion leaders (KOLs). In this paper, we aim to identify KOLs using artificial intelligence. We produce and explore a DCM co-authorship network, to characterise researchers' impact within the research field.

Methods

Through a bibliometric analysis of 1674 scientific papers in the DCM field, a co-authorship network was created. For each author, statistics about their connections to the co-authorship network (and so the nature of their collaboration) were generated. Using these connectedness statistics, a neural network was used to predict H-Index for each author (as a proxy for research impact). The neural network was retrospectively validated on an unseen author set.

Results

DCM research is regionally clustered, with strong collaboration across some international borders (e.g., North America) but not others (e.g., Western Europe). In retrospective validation, the neural network achieves a correlation coefficient of 0.86 (p<0.0001) between the true and predicted H-Index of each author. Thus, author impact can be accurately predicted using only the nature of an author's collaborations.

Discussion

Analysis of the neural network shows that the nature of collaboration strongly impacts an author's research visibility, and therefore suitability as a KOL. This also suggests greater collaboration within the DCM field could help to improve both individual research visibility and global synergy.

SUBMITTER: Grodzinski N

PROVIDER: S-EPMC8412256 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Degenerative cervical myelopathy (DCM) is a spinal cord condition that results in progressive non-traumatic compression of the cervical spinal cord. Spine surgeons must consider a large quantity of information relating to disease presentation, imaging features, and patient characteristics to determine if a patient will benefit from surgery for DCM. We applied a supervised machine learning approach to develop a classification model to predict individual patient outcome after surgery for DCM. Patients undergoing surgery for DCM as a part of the AOSpine CSM-NA or CSM-I prospective, multi-centre studies were included in the analysis. Out of 757 patients 605, 583, and 539 patients had complete follow-up information at 6, 12, and 24 months respectively and were included in the analysis. The primary outcome was improvement in the SF-6D quality of life indicator score by the minimum clinically important difference (MCID). The secondary outcome was improvement in the modified Japanese Orthopedic Association (mJOA) score by the MCID. Predictor variables reflected information about pre-operative disease severity, disease presentation, patient demographics, and comorbidities. A machine learning approach of feature engineering, data pre-processing, and model optimization was used to create the most accurate predictive model of outcome after surgery for DCM. Following data pre-processing 48, 108, and 101 features were chosen for model training at 6, 12, and 24 months respectively. The best performing predictive model used a random forest structure and had an average area under the curve (AUC) of 0.70, classification accuracy of 77%, and sensitivity of 78% when evaluated on a testing cohort that was not used for model training. Worse pre-operative disease severity, longer duration of DCM symptoms, older age, higher body weight, and current smoking status were associated with worse surgical outcomes. We developed a model that predicted positive surgical outcome for DCM with good accuracy at the individual patient level on an independent testing cohort. Our analysis demonstrates the applicability of machine-learning to predictive modeling in spine surgery.

Project description:BackgroundDegenerative cervical myelopathy (DCM) is widely accepted as the most common cause of adult myelopathy worldwide. Despite this, there is no specific term or diagnostic criteria in the International Classification of Diseases 11th Revision and no Medical Subject Headings (MeSH) or an equivalent in common literature databases. This makes searching the literature and thus conducting systematic reviews or meta-analyses imprecise and inefficient. Efficient research synthesis is integral to delivering evidence-based medicine and improving research efficiency.ObjectiveThis study aimed to illustrate the difficulties encountered when attempting to carry out a comprehensive and accurate evidence search in the field of DCM by identifying the key sources of imprecision and quantifying their impact.MethodsTo identify the key sources of imprecision and quantify their impact, an illustrative search strategy was developed using a validated DCM hedge combined with contemporary strategies used by authors in previous systematic reviews and meta-analyses. This strategy was applied to Medical Literature Analysis and Retrieval System Online (MEDLINE) and Excerpta Medica dataBASE (EMBASE) databases looking for relevant DCM systematic reviews and meta-analyses published within the last 5 years.ResultsThe MEDLINE via PubMed search strategy returned 24,166 results, refined to 534 papers after the application of inclusion and exclusion criteria. Of these, 32.96% (176/534) results were about DCM, and 18.16% (97/534) of these were DCM systematic reviews or meta-analyses. Non-DCM results were organized into imprecision categories (spinal: 268/534, 50.2%; nonspinal: 84/534, 15.5%; and nonhuman: 8/534, 1.5%). The largest categories were spinal cord injury (75/534, 13.67%), spinal neoplasms (44/534, 8.24%), infectious diseases of the spine and central nervous system (18/534, 3.37%), and other spinal levels (ie, thoracic, lumbar, and sacral; 18/534, 3.37%). Counterintuitively, the use of human and adult PubMed filters was found to exclude a large number of relevant articles. Searching a second database (EMBASE) added an extra 12 DCM systematic reviews or meta-analyses.ConclusionsDCM search strategies face significant imprecision, principally because of overlapping and heterogenous search terms, and inaccurate article indexing. Notably, commonly employed MEDLINE filters, human and adult, reduced search sensitivity, whereas the related articles function and the use of a second database (EMBASE) improved it. Development of a MeSH labeling and a standardized DCM definition would allow comprehensive and specific indexing of DCM literature. This is required to support a more efficient research synthesis.

Project description:BackgroundDegenerative cervical myelopathy (DCM), a progressive spinal cord injury caused by spinal cord compression from degenerative pathology, often presents with neck pain, sensorimotor dysfunction in the upper or lower limbs, gait disturbance, and bladder or bowel dysfunction. Its symptomatology is very heterogeneous, making early detection as well as the measurement or understanding of the underlying factors and their consequences challenging. Increasingly, evidence suggests that DCM may consist of subgroups of the disease, which are yet to be defined.ObjectiveThis study aimed to explore whether machine learning can identify clinically meaningful groups of patients based solely on clinical features.MethodsA survey was conducted wherein participants were asked to specify the clinical features they had experienced, their principal presenting complaint, and time to diagnosis as well as demographic information, including disease severity, age, and sex. K-means clustering was used to divide respondents into clusters according to their clinical features using the Euclidean distance measure and the Hartigan-Wong algorithm. The clinical significance of groups was subsequently explored by comparing their time to presentation, time with disease severity, and other demographics.ResultsAfter a review of both ancillary and cluster data, it was determined by consensus that the optimal number of DCM response groups was 3. In Cluster 1, there were 40 respondents, and the ratio of male to female participants was 13:21. In Cluster 2, there were 92 respondents, with a male to female participant ratio of 27:65. Cluster 3 had 57 respondents, with a male to female participant ratio of 9:48. A total of 6 people did not report biological sex in Cluster 1. The mean age in this Cluster was 56.2 (SD 10.5) years; in Cluster 2, it was 54.7 (SD 9.63) years; and in Cluster 3, it was 51.8 (SD 8.4) years. Patients across clusters significantly differed in the total number of clinical features reported, with more clinical features in Cluster 3 and the least clinical features in Cluster 1 (Kruskal-Wallis rank sum test: χ22=159.46; P<.001). There was no relationship between the pattern of clinical features and severity. There were also no differences between clusters regarding time since diagnosis and time with DCM.ConclusionsUsing machine learning and patient-reported experience, 3 groups of patients with DCM were defined, which were different in the number of clinical features but not in the severity of DCM or time with DCM. Although a clearer biological basis for the clusters may have been missed, the findings are consistent with the emerging observation that DCM is a heterogeneous disease, difficult to diagnose or stratify. There is a place for machine learning methods to efficiently assist with pattern recognition. However, the challenge lies in creating quality data sets necessary to derive benefit from such approaches.