Another Pre-Print
Samuel Håkansson, a postdoctoral researcher at our lab, together with Ph.D student Miklovana Tuci and our Group Lead of Paediatric Oncology Applications Sarah Brüningk have published a pre-print of the paper "Data-driven predictions of spinal cord injury: an exploration of current status and future prespectives".
external page 𝕃𝕚𝕟𝕜 𝕥𝕠 𝕥𝕙𝕖 𝕡𝕣𝕖𝕡𝕣𝕚𝕟𝕥.
In their paper, the team systematically reviewed the current literature on data-driven models of Spinal Cord Injury (hashtag#SCI) recovery prediction. SCI recovery prediction can be improved using machine learning (ML) approaches, which currently rely on commonly available data with challenges such as limited, missing, longitudinal data, and traditional ML models (linear regression and tree-based approaches). Despite increasing interest, the integration of these models into clinical practice requires advancements in handling limited and longitudinal data, ensuring model validity, and improving robustness. Achieving clinical translation necessitates careful consideration of model validity, causality, and practical deployability, along with the development and maintenance of increasingly detailed databases. Future research should focus on modern training paradigms and diverse data types to enhance the accuracy and applicability of SCI recovery predictions, ultimately benefiting rehabilitation strategies and clinical outcomes.
We congratulate Samuel, external page Mikla, Sarah, and colleagues for this great contribution.