Morphometric measures derived from spinal cord segmentations can serve as diagnostic and prognostic biomarkers in neurological diseases and injuries affecting the spinal cord. For instance, the spinal cord cross-sectional area can be used to monitor cord atrophy in multiple sclerosis and to characterize compression in degenerative cervical myelopathy. While robust, automatic segmentation methods to a wide variety of contrasts and pathologies have been developed over the past few years, whether their predictions are stable as the model is updated using new datasets has not been assessed. This is particularly important for deriving normative values from healthy participants. In this study, we present a spinal cord segmentation model trained on a multisite (n = 75 sites, 1,631 participants) dataset, including 9 different MRI contrasts and several spinal cord pathologies. We also introduce a lifelong learning framework to automatically monitor the morphometric drift as the model is updated using additional datasets. The framework is triggered by an automatic GitHub Actions workflow every time a new model is created, recording the morphometric values derived from the model’s predictions over time. As a real-world application of the proposed framework, we employed the spinal cord segmentation model to update a recently introduced normative database of healthy participants containing commonly used measures of spinal cord morphometry. Results showed that (i) our model performs well compared with its previous versions and existing pathology-specific models on the lumbar spinal cord, images with severe compression, and in the presence of intramedullary lesions and/or atrophy achieving an average Dice score of 0.95 ± 0.03; (ii) the automatic workflow for monitoring morphometric drift provides a quick feedback loop for developing future segmentation models; and (iii) the scaling factor required to update the database of morphometric measures is nearly constant among slices across the given vertebral levels, showing minimum drift between the current and previous versions of the model monitored by the framework. The code and model are open source and accessible via Spinal Cord Toolbox v7.0.
Canada lacks a national data source on individuals with limb loss and limb difference (LLD), which limits understanding of incidence, prevalence, risk factors, etiology, and healthcare outcomes. In the absence of standardized data collection, the provision of LLD healthcare services in Canada remains inconsistent. The objective of this study was to gather key interest groups’ perspectives on the development of a Canadian LLD registry. Invitees were identified through professional networks and snowball recruitment techniques. A two-round modified Delphi approach was utilized to identify key LLD registry domains via a pre-meeting survey followed by a virtual workshop on February 14, 2024. Of 96 invitees, 53 completed the survey, and 64 attended the workshop (63 from Canada and 1 from the United States). Five key LLD registry domains were identified: representation, standardization, practice-based evidence, research and innovation, and policy and funding. Inclusivity of diverse populations, national outcome measures adoption, and integration of psychosocial and clinical data was emphasized. Foreseen challenges included privacy concerns, necessary infrastructure, and resources to ensure long-term sustainability. Despite these challenges, a Canadian LLD registry could support advocacy, strengthen practice-based evidence, enhance research collaboration, improve clinical care, and inform population-level policies. Efforts to develop the registry are ongoing. Layman's Abstract Canada currently lacks a national registry of individuals with limb loss and limb difference (LLD), which limits our understanding of how many Canadians have limb loss or limb difference, and the effectiveness of healthcare services. Without consistent national data collection, it is impossible to understand regional differences in care, including prosthetic interventions and rehabilitation. This initiative gathered insights from representatives of the LLD community on establishing a Canadian LLD registry to improve research, healthcare delivery, and advocacy. Ninety-six LLD experts and professionals were identified to participate in a two-stage process that included a survey and virtual workshop. Fifty-three individuals (55%) completed the survey. On February 14, 2024, 64 (67%) individuals (63 Canadian and 1 American) attended the virtual workshop. The workshop included presentations on the development of the American Limb Loss and Preservation registry and the Canadian Rick Hansen Spinal Cord Injury registry, and measuring Canadian LLD outcomes. The meeting explored the feasibility, vision, and strategies for creating a Canadian LLD registry. Findings from the survey and workshop showed strong support for the registry, and workshop attendees discussed a number of critical areas for its development and sustainability. While some challenges regarding privacy, infrastructure, and sustainability still need to be addressed, there was agreement among meeting attendees on a strategy to start planning a registry. A Canadian registry would support advocacy efforts, enhance clinical care, encourage collaborative research, and provide essential data about care nationwide to ensure that the health and social care needs of the LLD community are met. Work is continuing to create and fund a Canadian LLD registry. Article PDF Link: https://jps.library.utoronto.ca/index.php/cpoj/article/view/46909/34865 How To Cite: Mayo A.L, Hitzig S.L, Zidarov D, MacKay C, Kaufman K.R, Noonan V.K, et al. A national strategy for a Canadian limb loss and limb difference registry. Canadian Prosthetics & Orthotics Journal. 2026; Volume 9, Issue 1, No. 3. https://doi.org/10.33137/cpoj.v9i1.46909 Corresponding Author: Dr. Amanda L. Mayo, MD, MHSc, FRCPC, Affiliations: 1) St. John’s Rehab, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 2) St. John’s Rehab Research Program, Sunnybrook Research Institute, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 3) Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada. E-Mail: amanda.mayo@sunnybrook.ca ORCID ID: https://orcid.org/0000-0001-7061-2529
Modern methods of data science have become powerful tools for analyzing complex, heterogeneous biomedical datasets, including those specific to spinal cord injury (SCI), enabling personalized predictions of recovery, identification of prognostic biomarkers, and insights into functional outcomes. This article summarizes the one-day Data Science Precourse, held at the 2025 annual scientific meeting of the American Spinal Injury Association (ASIA). The course was designed to illustrate advances in data science and their application to SCI research, while analyzing the unique challenges posed by SCI-specific data, such as sparse longitudinal measurements, variable injury characteristics, and diverse clinical and functional assessments. The precourse combined expert-led discussions with hands-on learning opportunities tailored for both clinical and data science audiences. Topics included addressing key challenges of artificial intelligence methods in SCI data analyses, such as learning from limited or incomplete datasets, implementing causal frameworks to understand recovery mechanisms, and ensuring robust and interpretable predictions for clinical decision-making. The event also showcased real-world applications of data science in SCI research, highlighting both solved and ongoing problems, including prognostic modeling, patient stratification, lesion analysis, and optimization of clinical trial design. The precourse culminated in the presentation by the winning teams of the 2025 ASIA Data Science Challenge.
Background: Establishing research partnerships can help close the research-practice gap. The Integrated Knowledge Translation Guiding Principles were developed in a spinal cord injury research context as a resource to facilitate research partnerships. Research funders play a significant role in the spinal cord injury research system. However, how funders define, require, evaluate, support research partnerships is rarely reported. This study identified spinal cord injury research funders in Canada and the United States, identified their approaches to supporting research partnerships, and explored organizational perspectives of principles of partnership. Methods: An environmental scan was conducted through five steps: (1) identifying spinal cord injury research organizations that funded the greatest number of spinal cord injury research publications in Canada and the United States between 2017 and 2022; (2) identifying one funding program related to research partnerships of each funder; (3) extracting online information of the programs; (4) interviewing funder informants; and (5) descriptive and deductive content analysis. The five steps were completed between April 2022 and September 2024. An additional data collection was conducted in July 2025 on a relevant National Institutes of Health funding program. Results: Sixteen organizations and seventeen partnership-supportive programs were identified. Six programs defined partnerships as researchers and research users engaging throughout the research process. Eleven programs required applicants to describe the partnership in applications and explicitly stated their peer review evaluation criteria. The programs supported research partnerships through remuneration for partners' engagement (n = 6), facilitating connections between researchers and potential partners (n = 3), and helping applicants prepare applications (n = 4). The programs had few strategies to evaluate awarded partnerships post-grant. Three descriptive categories emerged from the interviews: (1) Varied support for research partnerships; (2) Minimal capacity for partnership evaluation post-grant; and (3) Need for tools and resources to further support research partnerships. Conclusion: Differences existed in how spinal cord injury research funders in Canada and the United States defined, required, evaluated, and supported research partnerships. The results provided an initial landscape of funders' role in the spinal cord injury research system and may inform strategic efforts to optimizing meaningful engagement in a broader health research context.
Access to high-quality health information (HI) is critical for everyone involved in the research and management of medical conditions such as spinal cord injury (SCI). Recently, the use of Large Language Models (LLMs) through AI-based chatbots like ChatGPT has become increasingly integral to how people seek and consume HI. While LLMs have been evaluated in various clinical and health domains, there remains a notable gap in the literature regarding their use for SCI-specific questions. We conducted a narrative synthesis to identify the opportunities, challenges, and risks of using LLMs in SCI-related HI tasks, and to provide future direction for researchers, clinicians, and policymakers to better understand this fast-evolving landscape. We searched PubMed, Embase, and Google Scholar up to December 2025 and identified nine primary articles that investigated LLMs in the context of SCI-related queries. Our synthesis of the literature revealed that although there are promising results, these should be taken with caution due to mixed evidence for LLM's capability to effectively answer SCI-related questions. In addition, the LLM outputs were challenging to read, typically requiring an education level equivalent to a college-level student (grades 14-15) to be adequately understood. We recognize that LLMs can serve as valuable tools for accessing HI in SCI. However, LLMs can also pose significant risks, including the spread of mis- or dis-information that may be inaccurate or even dangerous, which can mislead individuals and caregivers, potentially resulting in detrimental health outcomes. Finally, methodological rigour needs to be improved to produce higher levels of evidence.
