Multi‐contrast unbiased MRI template of a Parkinson's disease population. (20th December 2022)

• Record Type: Journal Article
• Authors: Madge, Victoria; Fonov, Vladimir S; Xiao, Yiming; Zou, Lucy; Jackson, Courtney; Postuma, Ronald B; Dagher, Alain; Fon, Edward A; Collins, D Louis
• Is Part Of: Alzheimer's & dementia. Volume 18(2022)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-12-20
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.066612
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Parkinson's disease (PD) is a complex neurodegenerative disorder affecting regions such as the substantia nigra (SN), red nucleus (RN) and locus coeruleus (LC). Processing MRI data from patients with PD requires anatomical structural references for spatial normalization and structural segmentation. Extending our previous work (Xiao et al., 2015/17), we present multi‐contrast unbiased MRI templates of a larger population of 172 subjects using more modalities: T1w, T2 * w, T1‐T2 * fusion, R2 * map, T2w, PDw, fluid‐attenuated inversion recovery (FLAIR), improved susceptibility‐weighted imaging (CLEAR‐SWI) (Eckstein et al., 2021), and neuromelanin‐sensitive MRI (NM). Methods: Subjects were recruited through the Quebec Parkinson's Network since December 2018 and are included in a database of PD patients, RBD patients, and age‐ and sex‐ matched healthy controls, complete with MRI data, motor, and cognitive evaluation scores. All raw MRI was QC'd, eliminating scans with artefacts, motion, or noise, resulting in 172 subjects (69 female; ages=39‐87). Using our previous methods (Fonov et al., 2011), a non‐linear unbiased average T1‐T2 * fusion template was created in stereotaxic space. For the other contrasts, the subject's linear contrast‐to‐T1‐T2 * transformation was concatenated with the subject's T1‐T2 * non‐linear stereotaxic transformation to avoid multiple resamplings before voxel‐wise averaging of all subject data in the common template space. Results: Figure 1 depicts the whole brain 1mm template. Figure 2 depicts a closeup of the 0.3 mm template of the midbrain. From both figures, the first eight contrasts were created from 172 subjects: 145 PD, 1 RBD, and 26 controls. The NM contrast was created from 121 subjects (59 female; ages=39‐87): 99 PD, 1 RBD, and 21 controls. The T1‐T2 * fusion maps provide improved deep grey matter contrast from T2 * w images, while maintaining cortical structure information from T1w images. R2 * maps, T2w, PD2, FLAIR, and CLEAR‐SWI contrasts highlight deep grey structures including SN, RN, subthalamic nucleus (STN), putamen, and globus pallidum, while the NM contrast highlights NM‐rich regions of SN and LC. Conclusion: Anatomical structural references are useful for structural segmentation and morphometric analyses between healthy controls, PD, and RBD patients in regions such as SN, RN, and LC to aid in early diagnosis and better understand disease prognosis.