MSValid Data Collection

Dataset Composition and Patient Information

The dataset comprises MRI scans from 84 patients diagnosed with relapsing-remitting multiple sclerosis (RRMS), with each patient's data thoroughly anonymized to protect privacy.  The dataset includes a mixture of patients, some of whom have only a single MRI scan available, while others have longitudinal data with multiple scans taken at various timepoints. The intervals between these timepoints vary and, in certain cases, include instances where patients have undergone therapeutic interventions between scans. This allows for the examination of disease progression and treatment effects over time.

Individual patient data have been collected by Garibaldi Hospital (Catania) upon authorization of ethic committee COMITATO ETICO CATANIA 2 (prot. 810/C.E. del 20/12/2020, favorable opinion) and are irreversibly anonymised.

MRI Sequences and Imaging Details

Each MRI scan in the dataset consists of three primary sequences: Fluid Attenuated Inversion Recovery (FLAIR), T1-weighted, and T2-weighted imaging.

•       T1-weighted sequence: In this imaging modality, the white matter exhibits a hyperintense (lighter) signal compared to gray matter, which appears with a relatively lower intensity. The cerebrospinal fluid (CSF) is notably hypointense, presenting as a dark region within the brain’s ventricular and subarachnoid spaces.

•       T2-weighted sequence: Contrastingly, T2-weighted imaging depicts white matter as hypointense (darker), while gray matter appears with a higher signal intensity. CSF in this sequence is hyperintense, providing a bright signal, making it distinct in the brain’s fluid-filled spaces.

•       FLAIR sequence: The FLAIR imaging technique is similar to T2-weighted imaging but incorporates an inversion recovery pulse to suppress the CSF signal. This suppression enhances the visibility of lesions adjacent to the CSF, which might otherwise be obscured due to the high intensity of the fluid in standard T2-weighted images.

MS Lesion Characteristics and Detection

MS lesions exhibit distinct signal intensities across the different imaging sequences. On T1-weighted images, these lesions typically appear as hypointense areas, reflecting their lower signal intensity relative to normal white matter. Conversely, MS lesions are hyperintense on T2-weighted, proton density (PD)-weighted, and FLAIR sequences, indicating a higher signal intensity compared to surrounding brain tissues.

FLAIR images, in particular, provide superior lesion detection capabilities. The suppression of the CSF signal in FLAIR images enhances the contrast between lesions and surrounding tissues, making the lesions appear prominently hyperintense. This heightened contrast facilitates the precise delineation of MS lesions, which are often well-differentiated from adjacent brain structures in these images. The detailed imaging characteristics captured across these sequences are crucial for assessing the extent of MS-related brain damage, monitoring disease progression, and evaluating therapeutic outcomes.