Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease

• Published in: NeuroImage clinical Vol. 23; p. 101811
• Main Authors: Kim, Jun Pyo, Kim, Jeonghun, Park, Yu Hyun, Park, Seong Beom, Lee, Jin San, Yoo, Sole, Kim, Eun-Joo, Kim, Hee Jin, Na, Duk L., Brown, Jesse A., Lockhart, Samuel N., Seo, Sang Won, Seong, Joon-Kyung
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Inc 01.01.2019; Elsevier BV; Elsevier
• Subjects: Aged; Alzheimer Disease; Alzheimer Disease - classification; Alzheimer Disease - diagnostic imaging; Alzheimer Disease - pathology; Brain; Brain - diagnostic imaging; Brain - pathology; Classification model; Computer applications to medicine. Medical informatics; Diagnosis, Computer-Assisted; Diagnosis, Computer-Assisted - methods; Female; Frontotemporal Dementia; Frontotemporal Dementia - classification; Frontotemporal Dementia - diagnostic imaging; Frontotemporal Dementia - pathology; Humans; Machine Learning; Male; Middle Aged; Neurology. Diseases of the nervous system; Positron-Emission Tomography; R858-859.7; Radiology; RC346-429; Regular; Regular Article; Frontotemporal dementia; Classification model; Machine learning
• ISSN: 2213-1582; 2213-1582
• DOI: 10.1016/j.nicl.2019.101811

In a clinical setting, an individual subject classification model rather than a group analysis would be more informative. Specifically, the subtlety of cortical atrophy in some frontotemporal dementia (FTD) patients and overlapping patterns of atrophy among three FTD clinical syndromes including behavioral variant FTD (bvFTD), non-fluent/agrammatic variant primary progressive aphasia (nfvPPA), and semantic variant PPA (svPPA) give rise to the need for classification models at the individual level. In this study, we aimed to classify each individual subject into one of the diagnostic categories in a hierarchical manner by employing a machine learning-based classification method. We recruited 143 patients with FTD, 50 patients with Alzheimer's disease (AD) dementia, and 146 cognitively normal subjects. All subjects underwent a three-dimensional volumetric brain magnetic resonance imaging (MRI) scan, and cortical thickness was measured using FreeSurfer. We applied the Laplace Beltrami operator to reduce noise in the cortical thickness data and to reduce the dimension of the feature vector. Classifiers were constructed by applying both principal component analysis and linear discriminant analysis to the cortical thickness data. For the hierarchical classification, we trained four classifiers using different pairs of groups: Step 1 - CN vs. FTD + AD, Step 2 - FTD vs. AD, Step 3 - bvFTD vs. PPA, Step 4 - svPPA vs. nfvPPA. To evaluate the classification performance for each step, we used a10-fold cross-validation approach, performed 1000 times for reliability. The classification accuracy of the entire hierarchical classification tree was 75.8%, which was higher than that of the non-hierarchical classifier (73.0%). The classification accuracies of steps 1–4 were 86.1%, 90.8%, 86.9%, and 92.1%, respectively. Changes in the right frontotemporal area were critical for discriminating behavioral variant FTD from PPA. The left frontal lobe discriminated nfvPPA from svPPA, while the bilateral anterior temporal regions were critical for identifying svPPA. In the present study, our automated classifier successfully classified FTD clinical subtypes with good to excellent accuracy. Our classifier may help clinicians diagnose FTD subtypes with subtle cortical atrophy and facilitate appropriate specific interventions. •We developed a machine learning-based automated classifier for differential diagnosis of FTD clinical syndromes and AD.•Our classifier achieved good to excellent accuracy for each classification step.•Discriminative regions are similar to previously known cortical atrophic patterns in each clinical syndrome.