TY - JOUR
T1 - Generating diagnostic profiles of cognitive decline and dementia using magnetoencephalography
AU - Scheijbeler, Elliz P.
AU - Schoonhoven, Deborah N.
AU - Engels, Marjolein M. A.
AU - Scheltens, Philip
AU - Stam, Cornelis J.
AU - Gouw, Alida A.
AU - Hillebrand, Arjan
N1 - Funding Information: The authors would like to thank the participants of the Amsterdam Dementia Cohort for their contribution. Research of Amsterdam Alzheimer Center is part of the Neurodegeneration program of Amsterdam Neuroscience. The Amsterdam Alzheimer Center is supported by Alzheimer Nederland and Stichting VUmc funds. The authors thank technicians P.J. Ris, C.H. Plugge, N. Sijsma, N.C. Akemann, N. Zwagerman and M.C. Alting Siberg for acquisition of the MEG data. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Publisher Copyright: © 2021
PY - 2022/3
Y1 - 2022/3
N2 - Accurate identification of the underlying cause(s) of cognitive decline and dementia is challenging due to significant symptomatic overlap between subtypes. This study presents a multi-class classification framework for subjects with subjective cognitive decline, mild cognitive impairment, Alzheimer's disease, dementia with Lewy bodies, fronto-temporal dementia and cognitive decline due to psychiatric illness, trained on source-localized resting-state magnetoencephalography data. Diagnostic profiles, describing probability estimates for each of the 6 diagnoses, were assigned to individual subjects. A balanced accuracy rate of 41% and multi-class area under the curve value of 0.75 were obtained for 6-class classification. Classification primarily depended on posterior relative delta, theta and beta power and amplitude-based functional connectivity in the beta and gamma frequency band. Dementia with Lewy bodies (sensitivity: 100%, precision: 20%) and Alzheimer's disease subjects (sensitivity: 51%, precision: 90%) could be classified most accurately. Fronto-temporal dementia subjects (sensitivity: 11%, precision: 3%) were most frequently misclassified. Magnetoencephalography biomarkers hold promise to increase diagnostic accuracy in a noninvasive manner. Diagnostic profiles could provide an intuitive tool to clinicians and may facilitate implementation of the classifier in the memory clinic.
AB - Accurate identification of the underlying cause(s) of cognitive decline and dementia is challenging due to significant symptomatic overlap between subtypes. This study presents a multi-class classification framework for subjects with subjective cognitive decline, mild cognitive impairment, Alzheimer's disease, dementia with Lewy bodies, fronto-temporal dementia and cognitive decline due to psychiatric illness, trained on source-localized resting-state magnetoencephalography data. Diagnostic profiles, describing probability estimates for each of the 6 diagnoses, were assigned to individual subjects. A balanced accuracy rate of 41% and multi-class area under the curve value of 0.75 were obtained for 6-class classification. Classification primarily depended on posterior relative delta, theta and beta power and amplitude-based functional connectivity in the beta and gamma frequency band. Dementia with Lewy bodies (sensitivity: 100%, precision: 20%) and Alzheimer's disease subjects (sensitivity: 51%, precision: 90%) could be classified most accurately. Fronto-temporal dementia subjects (sensitivity: 11%, precision: 3%) were most frequently misclassified. Magnetoencephalography biomarkers hold promise to increase diagnostic accuracy in a noninvasive manner. Diagnostic profiles could provide an intuitive tool to clinicians and may facilitate implementation of the classifier in the memory clinic.
KW - Cognitive decline
KW - Dementia
KW - Machine learning
KW - Magnetoencephalography
KW - Multi-class classification
UR - http://www.scopus.com/inward/record.url?scp=85121099453&partnerID=8YFLogxK
U2 - https://doi.org/10.1016/j.neurobiolaging.2021.11.002
DO - https://doi.org/10.1016/j.neurobiolaging.2021.11.002
M3 - Article
C2 - 34906377
SN - 0197-4580
VL - 111
SP - 82
EP - 94
JO - Neurobiology of aging
JF - Neurobiology of aging
ER -