Multimodal MRI biomarkers optimize differentiation between progressive supranuclear palsy and Parkinson's disease

BACKGROUND: It remains a challenge to accurately differentiate progressive supranuclear palsy (PSP) and Parkinson's disease (PD). This study aimed to investigate the value of multimodal MRI biomarkers assessed by a single 3D multi-echo magnetization transfer contrast gradient echo (3D ME-MTC-GRE) sequence in the differential diagnosis of PSP from PD. METHODS: Seventy-three PD patients, 32 PSP patients, and 36 age - and sex-matched healthy controls were prospectively enrolled. MRI examinations were performed at a 3.0 T MRI scanner. The midbrain to pons ratio (MPR) was measured in the midsagittal T1-weighted imaging. Neuromelanin-sensitive MRI and quantitative susceptibility mapping were simultaneously obtained by a single 3D ME-MTC-GRE sequence. Substantia nigra pars compacta (SNpc) volume and the red nucleus (RN) susceptibility were analyzed and compared among three groups by a single 3D ME-MTC-GRE sequence. The diagnostic performance of these biomarkers was assessed by areas under curves (AUCs). RESULTS: The PSP group exhibited the smallest MPR and SNpc volume and the highest RN susceptibility compared with the other two groups. The PD group also showed a smaller SNpc volume than HC group. The MPR, SNpc volume, and RN susceptibility on their own differentiated PSP from PD with AUCs of 0.875, 0.883, and 0.892, respectively. The model consisting of SNpc volume and RN susceptibility, as well as the model consisting of these three biomarkers, demonstrated better diagnostic performance in distinguishing PSP from PD patients, with AUCs of 0.928 and 0.939, respectively. CONCLUSION: The multimodal MRI biomarkers offer a more comprehensive understanding of PSP and PD patients from both morphological and pathological perspectives. By combining these MRI biomarkers, it may have the potential to promote reliable MRI-based identification of PSP and PD patients, and facilitate more accurate clinical diagnosis.