A reliable and accurate quantification of brain tissue loss is important to measure progressive atrophy caused by neurological diseases such as multiple sclerosis. However, accuracy and reproducibility of current methods are often limited by partial volume effects, especially at tissue interfaces where subtle atrophy patterns are likely to occur. We propose a longitudinal pipeline for brain tissue segmentation incorporating partial volume estimation to increase longitudinal robustness. Results show an increase in reproducibility of 44% compared to methods not including partial volume effects in volume estimation, suggesting that these effects should be taken into account for longitudinal atrophy measurements.
This abstract and the presentation materials are available to members only; a login is required.