What is the Praxis AI Lab?
Smarter SCI research and care, powered by AI.
The Praxis Artificial Intelligence (AI) Lab uses advanced data analysis methods such as machine learning and deep learning to develop algorithms and clinical tools that improve SCI care by assessing risks for secondary complications and predicting recovery outcomes. Key applications of AI:
- Data Analysis: AI algorithms process large amount of data (numeric or medical images) to uncover patterns too complex for human analysis.
- Modeling and Simulation: AI creates predictive models and simulations of complex systems, like the healthcare system, to forecast outcomes more efficiently
- Decision Support: AI analyzes patient data to assist clinicians in making accurate diagnoses and treatment decisions.
Why is Praxis Using AI?
Data-driven answers for SCI recovery.
Clinicians and researchers often ask: How do age and injury severity affect survival after traumatic SCI? How likely is independent walking?
With large datasets like the Praxis-led Canadian National SCI Registry (RHSCIR), AI can provide data-driven answers using predictive algorithms and clinical prediction rules (CPRs). These algorithms mimic human decision-making, leveraging neural networks that learn from data to improve outcome predictions.
AI and machine learning are also transforming SCI research design. Techniques such as Decision Trees and Unbiased Recursive Partitioning allow researchers to group patients into smaller, more homogeneous categories based on injury and clinical profiles – paying the way for personalized treatment strategies.
Further reading on how Praxis is using Decision Trees and Unbiased Recursive Partitioning in SCI research.
How is Praxis Advancing AI in Research & Care?
Turning data into predictive tools for clinicians and patients.
Using the wealth of data in our Registry and collaborating with other AI experts, Praxis AI Lab is:
- Developing predictive algorithms for SCI outcomes:
- In-hospital and 1-year mortality – using the SCI Risk Score (SCIRS) (Fallah 2022)
- Long-term survival – using body mass index (BMI) in classification and regression tree (CART) analysis and generalized additive models (Fallah 2024)
- Health care utilization and health outcomes – using the multi-morbidity index (MMI) of 30 secondary health conditions (SHCs) in traumatic (Noonan 2014) and non-traumatic SCI (Hong 2023), or a network-derived MMI, consisting of 25 SHCs (Fallah 2024)
- Making predictive algorithms user-friendly for clinicians and PLEX:
- Ambulation App – a web-based tool to predict the chance of walking independently 1-year after traumatic SCI (Hakimjavadi 2023)
- Leading global research:
- Special Issue in Frontiers in Neurology – Over 100,000 views! Articles in this issue focus on epidemiology of SCI and the strategic use of AI in neurological recovery and clinical measurements in SCI.
- Due to its high impact and broad interest, this issue has been compiled into an e-book edited by Dr. Nader Fallah (Associate Director of Praxis AI Lab) and Dr. Vanessa Noonan (Director of Research & Care from Praxis) and Dr. Lisa Sharwood (University of New South Wales in Australia).
What is Next?
Personalized care and insights for every person’s journey.
The Praxis AI Lab is expanding AI applications to:
- Non-traumatic SCI – particularly in degenerative cervical myelopathy (DCM), an age-related spinal degeneration and the most common cause of spinal cord dysfunction in adults worldwide.
- Imaging data – leveraging MRI data from national Canadian SCI Imaging Repository to improve prognosis and predict recovery in traumatic SCI.