Translating state-of-the-art brain magnetic resonance imaging (MRI) techniques into clinical practice: multimodal MRI differentiates dementia subtypes in a traditional clinical setting

• Published in: Quantitative imaging in medicine and surgery Vol. 11; no. 9; pp. 4056 - 4073
• Main Authors: Kuhn, Taylor, Becerra, Sergio, Duncan, John, Spivak, Norman, Dang, Bianca Huan, Habelhah, Barshen, Mahdavi, Kennedy D, Mamoun, Michael, Whitney, Michael, Pereles, F Scott, Bystritsky, Alexander, Jordan, Sheldon E
• Format: Journal Article
• Language: English
• Published: China AME Publishing Company 01-09-2021
• Subjects: Original; Magnetic resonance imaging (MRI); neurodegenerative; dementia; Alzheimer’s; Parkinson’s
• ISSN: 2223-4292; 2223-4306
• DOI: 10.21037/qims-20-1355

This study sought to validate the clinical utility of multimodal magnetic resonance imaging (MRI) techniques in the assessment of neurodegenerative disorders. We intended to demonstrate that advanced neuroimaging techniques commonly used in research can effectively be employed in clinical practice to accurately differentiate heathy aging and dementia subtypes.
Twenty patients with dementia of the Alzheimer's type (DAT) and 18 patients with Parkinson's disease dementia (PDD) were identified using gold-standard techniques. Twenty-three healthy, age and sex matched control participants were also recruited. All participants underwent multimodal MRI including T1 structural, diffusion tensor imaging (DTI), arterial spin labeling (ASL), and magnetic resonance spectroscopy (MRS). MRI modalities were evaluated by trained neuroimaging readers and were separately assessed using cross-validated, iterative discriminant function analyses with subsequent feature reduction techniques. In this way, each modality was evaluated for its ability to differentiate patients with dementia from healthy controls as well as to differentiate dementia subtypes.
Following individual and group feature reduction, each of the multimodal MRI metrics except MRS successfully differentiated healthy aging from dementia and also demonstrated distinct dementia subtypes. Using the following ten metrics, excellent separation (95.5% accuracy, 92.3% sensitivity; 100.0% specificity) was achieved between healthy aging and neurodegenerative conditions: volume of the left frontal pole, left occipital pole, right posterior superior temporal gyrus, left posterior cingulate gyrus, right planum temporale; perfusion of the left hippocampus and left occipital lobe; fractional anisotropy (FA) of the forceps major and bilateral anterior thalamic radiation. Using volume of the left frontal pole, right posterior superior temporal gyrus, left posterior cingulate gyrus, perfusion of the left hippocampus and left occipital lobe; FA of the forceps major and bilateral anterior thalamic radiation, neurodegenerative subtypes were accurately differentiated as well (87.8% accuracy, 95.2% sensitivity; 85.0% specificity).
Regional volumetrics, DTI metrics, and ASL successfully differentiated dementia patients from controls with sufficient sensitivity to differentiate dementia subtypes. Similarly, feature reduction results suggest that advanced analyses can meaningfully identify brain regions with the most positive predictive value and discriminant validity. Together, these advanced neuroimaging techniques can contribute significantly to diagnosis and treatment planning for individual patients.