1
|
Khan YF, Kaushik B, Chowdhary CL, Srivastava G. Ensemble Model for Diagnostic Classification of Alzheimer's Disease Based on Brain Anatomical Magnetic Resonance Imaging. Diagnostics (Basel) 2022; 12:diagnostics12123193. [PMID: 36553199 PMCID: PMC9777931 DOI: 10.3390/diagnostics12123193] [Citation(s) in RCA: 8] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/27/2022] [Revised: 11/08/2022] [Accepted: 11/15/2022] [Indexed: 12/24/2022] Open
Abstract
Alzheimer's is one of the fast-growing diseases among people worldwide leading to brain atrophy. Neuroimaging reveals extensive information about the brain's anatomy and enables the identification of diagnostic features. Artificial intelligence (AI) in neuroimaging has the potential to significantly enhance the treatment process for Alzheimer's disease (AD). The objective of this study is two-fold: (1) to compare existing Machine Learning (ML) algorithms for the classification of AD. (2) To propose an effective ensemble-based model for the same and to perform its comparative analysis. In this study, data from the Alzheimer's Diseases Neuroimaging Initiative (ADNI), an online repository, is utilized for experimentation consisting of 2125 neuroimages of Alzheimer's disease (n = 975), mild cognitive impairment (n = 538) and cognitive normal (n = 612). For classification, the framework incorporates a Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB), and K-Nearest Neighbor (K-NN) followed by some variations of Support Vector Machine (SVM), such as SVM (RBF kernel), SVM (Polynomial Kernel), and SVM (Sigmoid kernel), as well as Gradient Boost (GB), Extreme Gradient Boosting (XGB) and Multi-layer Perceptron Neural Network (MLP-NN). Afterwards, an Ensemble Based Generic Kernel is presented where Master-Slave architecture is combined to attain better performance. The proposed model is an ensemble of Extreme Gradient Boosting, Decision Tree and SVM_Polynomial kernel (XGB + DT + SVM). At last, the proposed method is evaluated using cross-validation using statistical techniques along with other ML models. The presented ensemble model (XGB + DT + SVM) outperformed existing state-of-the-art algorithms with an accuracy of 89.77%. The efficiency of all the models was optimized using Grid-based tuning, and the results obtained after such process showed significant improvement. XGB + DT + SVM with optimized parameters outperformed all other models with an efficiency of 95.75%. The implication of the proposed ensemble-based learning approach clearly shows the best results compared to other ML models. This experimental comparative analysis improved understanding of the above-defined methods and enhanced their scope and significance in the early detection of Alzheimer's disease.
Collapse
Affiliation(s)
| | - Baijnath Kaushik
- School of CSE, Shri Mata Vaishno Devi University, Katra 182320, India
| | - Chiranji Lal Chowdhary
- School of Information Technology and Engineering, Vellore Institute of Technology, Vellore 632014, India
- Correspondence:
| | - Gautam Srivastava
- Department of Mathematics and Computer Science, Brandon University, Brandon, MB R7A 6A9, Canada
- Research Centre for Interneural Computing, China Medical University, Taichung 40402, Taiwan
- Department of Computer Science and Math, Lebanese American University, Beirut 1102, Lebanon
| |
Collapse
|
2
|
Townley RA, Polsinelli AJ, Fields JA, Machulda MM, Jones DT, Graff-Radford J, Kantarci KM, Lowe VJ, Rademakers RV, Baker MC, Kumar N, Boeve BF. Longitudinal clinical, neuropsychological, and neuroimaging characterization of a kindred with a 12-octapeptide repeat insertion in PRNP: the next generation. Neurocase 2020; 26:211-219. [PMID: 32602775 PMCID: PMC7426006 DOI: 10.1080/13554794.2020.1787458] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 10/18/2019] [Accepted: 06/18/2020] [Indexed: 01/28/2023]
Abstract
BACKGROUND Highly penetrant inherited mutations in the prion protein gene (PRNP) offer a window to study the pathobiology of prion disorders. METHOD Clinical, neuropsychological, and neuroimaging characterization of a kindred. RESULTS Three of four mutation carriers have progressed to a frontotemporal dementia phenotype. Declines in neuropsychological function coincided with changes in FDG-PET at the identified onset of cognitive impairment. CONCLUSIONS AND RELEVANCE Gene silencing treatments are on the horizon and when they become available, early detection will be crucial. Longitudinal studies involving familial mutation kindreds can offer important insights into the initial neuropsychological and neuroimaging changes necessary for early detection.
Collapse
Affiliation(s)
- Ryan A. Townley
- Department of Neurology, University of Kansas Medical Center, Kansas City, KS 66160
| | | | - Julie A. Fields
- Department of Psychiatry and Psychology, Mayo Clinic, Rochester, MN, USA 55902
| | - Mary M. Machulda
- Department of Psychiatry and Psychology, Mayo Clinic, Rochester, MN, USA 55902
| | - David T. Jones
- Department of Neurology, Indiana University School of Medicine, IN, USA 46202
- Department of Diagnostic Radiology, Mayo Clinic, Rochester, MN, USA 55902
| | | | - Kejal M. Kantarci
- Department of Diagnostic Radiology, Mayo Clinic, Rochester, MN, USA 55902
| | - Val J. Lowe
- Department of Diagnostic Radiology, Mayo Clinic, Rochester, MN, USA 55902
| | | | - Matt C. Baker
- Department of Neuroscience, Mayo Clinic, Jacksonville, FL, USA 32224
| | - Neeraj Kumar
- Department of Neurology, Indiana University School of Medicine, IN, USA 46202
| | - Bradley F. Boeve
- Department of Neurology, Indiana University School of Medicine, IN, USA 46202
| |
Collapse
|
3
|
Llibre-Guerra JJ, Li Y, Schindler SE, Gordon BA, Fagan AM, Morris JC, Benzinger TLS, Hassenstab J, Wang G, Allegri R, Berman SB, Chhatwal J, Farlow MR, Holtzman DM, Jucker M, Levin J, Noble JM, Salloway S, Schofield P, Karch C, Fox NC, Xiong C, Bateman RJ, McDade E. Association of Longitudinal Changes in Cerebrospinal Fluid Total Tau and Phosphorylated Tau 181 and Brain Atrophy With Disease Progression in Patients With Alzheimer Disease. JAMA Netw Open 2019; 2:e1917126. [PMID: 31825500 PMCID: PMC6991202 DOI: 10.1001/jamanetworkopen.2019.17126] [Citation(s) in RCA: 17] [Impact Index Per Article: 3.4] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 12/16/2022] Open
Abstract
IMPORTANCE The amyloid/tau/neurodegeneration (A/T/N) framework uses cerebrospinal fluid (CSF) levels of total tau (tTau) as a marker of neurodegeneration and CSF levels of phosphorylated tau 181 (pTau181) as a marker of tau tangles. However, it is unclear whether CSF levels of tTau and pTau181 have similar or different trajectories over the course of Alzheimer disease. OBJECTIVES To examine the rates of change in CSF levels of tTau and pTau181 across the Alzheimer disease course and how the rates of change are associated with brain atrophy as measured by magnetic resonance imaging. DESIGN, SETTING, AND PARTICIPANTS This cohort study was set in tertiary research clinics. Each participant was a member of a pedigree with a known mutation for dominantly inherited Alzheimer disease. Participants were divided into 3 groups on the basis of the presence of a mutation and their Clinical Dementia Rating score. Data analysis was performed in June 2019. MAIN OUTCOMES AND MEASURES Rates of change of CSF tTau and pTau181 levels and their association with the rate of change of brain volume. RESULTS Data from 465 participants (283 mutation carriers and 182 noncarriers) were analyzed. The mean (SD) age of the cohort was 37.8 (11.3) years, and 262 (56.3%) were women. The mean (SD) follow-up duration was 2.7 (1.5) years. Two or more longitudinal CSF and magnetic resonance imaging assessments were available for 160 and 247 participants, respectively. Sixty-five percent of mutation carriers (183) did not have symptoms at baseline (Clinical Dementia Rating score, 0). For mutation carriers, the annual rates of change for CSF tTau and pTau181 became significantly different from 0 approximately 10 years before the estimated year of onset (mean [SE] rates of change, 5.5 [2.8] for tTau [P = .05] and 0.7 [0.3] for pTau 181 [P = .04]) and 15 years before onset (mean [SE] rates of change, 5.4 [3.9] for tTau [P = .17] and 1.1 [0.5] for pTau181 [P = .03]), respectively. The rate of change of pTau181 was positive and increased at the early stages of the disease, showing a positive rate of change starting at 15 estimated years before onset until 5 estimated years before onset (mean [SE], 0.4 [0.3]), followed by a positive but decreasing rate of change at year 0 (mean [SE], 0.1 [0.3]) and then negative rates of change at 5 years (mean [SE], -0.3 [0.4]) and 10 years (mean [SE], -0.6 [0.6]) after symptom onset. In individuals without symptoms (Clinical Dementia Rating score, 0), the rates of change of CSF tTau and pTau181 were negatively associated with brain atrophy (high rates of change in CSF measures were associated with low rates of change in brain volume in asymptomatic stages). After symptom onset (Clinical Dementia Rating score, >0), an increased rate of brain atrophy was not associated with rates of change of levels of both CSF tTau and pTau181. CONCLUSIONS AND RELEVANCE These findings suggest that CSF tTau and pTau181 may have different associations with brain atrophy across the disease time course. These results have implications for understanding the dynamics of disease pathobiology and interpreting neuronal injury biomarker concentrations in response to Alzheimer disease progression and disease-modifying therapies.
Collapse
Affiliation(s)
| | - Yan Li
- Department of Biostatistics, Washington University in St Louis, St Louis, Missouri
| | | | - Brian A. Gordon
- Department of Radiology, Washington University in St Louis, St Louis, Missouri
| | - Anne M. Fagan
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
- Department of Pathology and Immunology, Washington University School of Medicine in St Louis, St Louis, Missouri
| | - John C. Morris
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
- Hope Center for Neurological Disorders, St Louis, Missouri
- Knight Alzheimer’s Disease Research Center, St Louis, Missouri
| | | | - Jason Hassenstab
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
| | - Guoqiao Wang
- Hertie Institute for Clinical Brain Research, Department of Cellular Neurology, University of Tübingen, Tübingen, Germany
| | - Ricardo Allegri
- Department of Cognitive Neurology, Institute for Neurological Research Fleni, Buenos Aires, Argentina
| | - Sarah B. Berman
- Department of Radiology, Washington University in St Louis, St Louis, Missouri
- Department of Neurology, University of Pittsburgh, Pittsburgh, Pennsylvania
| | | | | | - David M. Holtzman
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
- Hope Center for Neurological Disorders, St Louis, Missouri
- Knight Alzheimer’s Disease Research Center, St Louis, Missouri
| | - Mathias Jucker
- Hertie Institute for Clinical Brain Research, Department of Cellular Neurology, University of Tübingen, Tübingen, Germany
- DZNE-German Center for Neurodegenerative Diseases, Tübingen, Tübingen, Germany
| | - Johannes Levin
- Department of Neurology, Ludwig-Maximilians-University, Munich, Germany
- DZNE-German Center for Neurodegenerative Diseases, Munich, Munich, Germany
- SyNergy, Munich Cluster for Systems Neurology, Munich, Germany
| | - James M. Noble
- Taub Institute for Research on Alzheimer’s Disease, Aging Brain G.H. Sergievsky Center, Department of Neurology, Columbia University Medical Center, New York, New York
| | - Stephen Salloway
- Memory & Aging Program, Butler Hospital, Brown University, Providence, Rhode Island
| | - Peter Schofield
- Neuroscience Research Australia, Randwick, Sydney, New South Wales, Australia
- School of Medical Sciences, UNSW Sydney, Sydney, New South Wales, Australia
| | - Celeste Karch
- Department of Psychiatry, Washington University in St Louis, St Louis, Missouri
| | - Nick C. Fox
- Dementia Research Centre, University College London, London, United Kingdom
| | - Chengjie Xiong
- Department of Biostatistics, Washington University in St Louis, St Louis, Missouri
| | - Randall J. Bateman
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
| | - Eric McDade
- Department of Neurology, Washington University in St Louis, St Louis, Missouri
| |
Collapse
|