In this challenge, participants will attempt to prognosticate long-term recovery after spinal cord injury by analyzing a subset of the data from the [Sygen multicenter clinical trial](https://pubmed.ncbi.nlm.nih.gov/11805614/). The data used to make the prediction of recovery at 6-12 months will be data available in the first month after spinal cord injury. One piece of data that is commonly used to make such predictions are the [International Standards for Neurological Classification of SCI (ISNCSCI)](https://asia-spinalinjury.org/international-standards-neurological-classification-sci-isncsci-worksheet/). This clinical outcome assessment evaluates motor and sensory impairment across the body and helps localize the neurologic level of the spinal cord injury and grade the severity of that injury. Additional data includes demographic information, comorbidities, surgical interventions, and medical complications.

This data challenge involves two tracks predicting different outcomes. These reflect two constructs within the [World Health Organization International Classification of Functioning, Disability and Health (ICF)](https://www.who.int/standards/classifications/international-classification-of-functioning-disability-and-health). [The first construct](https://www.kaggle.com/competitions/spinal-cord-injury-challenge-impairment-track) is future impairment of body structure and function, assessed through the ISNCSCI scores at 26-52 weeks after SCI. Impairments, such as strength and sensation, are not the same construct as real world function. **This track is to predict the quality of walking** on an 8-point modified Benzel classification, more frequently referred to as the lower extremity subscore of the [modified Japanese Orthopedic Association](https://pmc.ncbi.nlm.nih.gov/articles/PMC4383381/) score proposed by [Benzel](https://pubmed.ncbi.nlm.nih.gov/1802159/).

Why these outcomes? Essentially all functions of the body are influenced by spinal cord injury and are important to predict, including arm and hand function, bowel and bladder function, respiratory function, blood pressure control, and skin integrity. However, there are no available datasets that capture these rich aspects of function. Our goal with this challenge was to predict both neurological recovery (scored with the ISNCSI) and an example of real world function (walking scored with the Benzel classification).

A representative from the winning team for each track will be invited to present their work at the Data Science precourse at the ASIA Society annual meeting, with travel support available for the representatives.

There are additional open datasets available with data for spinal cord injury. Participants are invited to also use these data if it improves their prediction performance on the Sygen challenge data but need to provide information and links to any data that was used - teams using other datasets not publicly available, including private datasets from their health care system, will not be eligible for invitation to the ASIA Society data science precourse. Any teams with access to the complete Sygen dataset are prohibited from using this data when training their algorithms. Leading teams will be asked to submit a write-up and code to reproduce their solution.