We propose a method to estimate multiple sclerosis (MS) lesion age (less than or greater than a year old) using non-gadolinium magnetic resonance imaging. The method utilizes the less invasive Quantitative Susceptibility Map. Radiomic features are calculated over a lesion and a random forest classification model is used. In a validation set, the model has an AUC of 0.79 (95% CI: [0.63, 0.86]) and an accuracy of 0.73 (95% CI: [0.60, 0.80]). This method can be used to aid in the diagnosis of MS, as part of the diagnostic criteria is to show lesion dissemination in time.
This abstract and the presentation materials are available to members only; a login is required.