Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease
Background: 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 i...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2019-01-01
|
| Series: | NeuroImage: Clinical |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2213158219301615 |
| _version_ | 1818990220791513088 |
|---|---|
| author | Jun Pyo Kim Jeonghun Kim Yu Hyun Park Seong Beom Park Jin San Lee Sole Yoo Eun-Joo Kim Hee Jin Kim Duk L. Na Jesse A. Brown Samuel N. Lockhart Sang Won Seo Joon-Kyung Seong |
| author_facet | Jun Pyo Kim Jeonghun Kim Yu Hyun Park Seong Beom Park Jin San Lee Sole Yoo Eun-Joo Kim Hee Jin Kim Duk L. Na Jesse A. Brown Samuel N. Lockhart Sang Won Seo Joon-Kyung Seong |
| author_sort | Jun Pyo Kim |
| collection | DOAJ |
| description | Background: 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. Methods: 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. Results: 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. Conclusions: 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. Keywords: Frontotemporal dementia, Classification model, Machine learning |
| first_indexed | 2024-12-20T19:50:55Z |
| format | Article |
| id | doaj.art-9fbe2927595b48a4ba18fdc060874095 |
| institution | Directory Open Access Journal |
| issn | 2213-1582 |
| language | English |
| last_indexed | 2024-12-20T19:50:55Z |
| publishDate | 2019-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage: Clinical |
| spelling | doaj.art-9fbe2927595b48a4ba18fdc0608740952022-12-21T19:28:19ZengElsevierNeuroImage: Clinical2213-15822019-01-0123Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's diseaseJun Pyo Kim0Jeonghun Kim1Yu Hyun Park2Seong Beom Park3Jin San Lee4Sole Yoo5Eun-Joo Kim6Hee Jin Kim7Duk L. Na8Jesse A. Brown9Samuel N. Lockhart10Sang Won Seo11Joon-Kyung Seong12Department of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of KoreaDepartment of Bio-convergence Engineering, Korea University, Seoul, Republic of KoreaDepartment of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of KoreaDepartment of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of KoreaDepartment of Neurology, Kyunghee University Medical Center, Seoul, Republic of KoreaDepartment of Cognitive Science, Yonsei University, Seoul, Republic of KoreaDepartment of Neurology, Busan National University Hospital, Busan, Republic of KoreaDepartment of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of KoreaDepartment of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of KoreaDepartment of Neurology, UCSF Weill Institute for Neurosciences, University of California, San Francisco, San Francisco, CA, USADepartment of Internal Medicine, Section of Gerontology and Geriatric Medicine, Wake Forest School of Medicine, Winston-Salem, NC, USADepartment of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea; Neuroscience Center, Samsung Medical Center, Seoul, Republic of Korea; Samsung Alzheimer Research Center, Samsung Medical Center, Seoul, Republic of Korea; Center for Clinical Epidemiology, Samsung Medical Center, Seoul, Republic of Korea; Department of Clinical Research Design and Evaluation, SAIHST, Sungkyunkwan University, Seoul, Republic of Korea; Correspondence to: S. W. Seo, Department of Neurology, Samsung Medical Center, Sungkyunkwan University, School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Republic of Korea.Department of Bio-convergence Engineering, Korea University, Seoul, Republic of Korea; School of Biomedical Engineering, Korea University, Seoul, Republic of Korea; Correspondence to: J. K. Seong, School of Biomedical Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul, Republic of Korea.Background: 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. Methods: 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. Results: 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. Conclusions: 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. Keywords: Frontotemporal dementia, Classification model, Machine learninghttp://www.sciencedirect.com/science/article/pii/S2213158219301615 |
| spellingShingle | Jun Pyo Kim Jeonghun Kim Yu Hyun Park Seong Beom Park Jin San Lee Sole Yoo Eun-Joo Kim Hee Jin Kim Duk L. Na Jesse A. Brown Samuel N. Lockhart Sang Won Seo Joon-Kyung Seong Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease NeuroImage: Clinical |
| title | Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease |
| title_full | Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease |
| title_fullStr | Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease |
| title_full_unstemmed | Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease |
| title_short | Machine learning based hierarchical classification of frontotemporal dementia and Alzheimer's disease |
| title_sort | machine learning based hierarchical classification of frontotemporal dementia and alzheimer s disease |
| url | http://www.sciencedirect.com/science/article/pii/S2213158219301615 |
| work_keys_str_mv | AT junpyokim machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT jeonghunkim machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT yuhyunpark machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT seongbeompark machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT jinsanlee machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT soleyoo machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT eunjookim machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT heejinkim machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT duklna machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT jesseabrown machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT samuelnlockhart machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT sangwonseo machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease AT joonkyungseong machinelearningbasedhierarchicalclassificationoffrontotemporaldementiaandalzheimersdisease |