Morphometric measures derived from spinal cord segmentations can serve as diagnostic and prognostic biomarkers in neurological diseases and injuries affecting the spinal cord. 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 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 outperforms previous versions and pathology-specific models on challenging lumbar spinal cord cases, achieving an average Dice score of ; (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 model is freely available in Spinal Cord Toolbox v7.0.
View on arXiv@article{karthik2025_2505.01364,
title={ Monitoring morphometric drift in lifelong learning segmentation of the spinal cord },
author={ Enamundram Naga Karthik and Sandrine Bédard and Jan Valošek and Christoph S. Aigner and Elise Bannier and Josef Bednařík and Virginie Callot and Anna Combes and Armin Curt and Gergely David and Falk Eippert and Lynn Farner and Michael G Fehlings and Patrick Freund and Tobias Granberg and Cristina Granziera and RHSCIR Network Imaging Group and Ulrike Horn and Tomáš Horák and Suzanne Humphreys and Markus Hupp and Anne Kerbrat and Nawal Kinany and Shannon Kolind and Petr Kudlička and Anna Lebret and Lisa Eunyoung Lee and Caterina Mainero and Allan R. Martin and Megan McGrath and Govind Nair and Kristin P. O'Grady and Jiwon Oh and Russell Ouellette and Nikolai Pfender and Dario Pfyffer and Pierre-François Pradat and Alexandre Prat and Emanuele Pravatà and Daniel S. Reich and Ilaria Ricchi and Naama Rotem-Kohavi and Simon Schading-Sassenhausen and Maryam Seif and Andrew Smith and Seth A Smith and Grace Sweeney and Roger Tam and Anthony Traboulsee and Constantina Andrada Treaba and Charidimos Tsagkas and Zachary Vavasour and Dimitri Van De Ville and Kenneth Arnold Weber II and Sarath Chandar and Julien Cohen-Adad },
journal={arXiv preprint arXiv:2505.01364},
year={ 2025 }
}