Biological brain-age prediction using machine learning on neuroimaging data: links with pathophysiological mechanisms, dementia risk factors and cognitive decline

This thesis explores the links between neuroimaging-derived brain-age, which is computed using machine learning techniques, and different pathophysiological mechanisms, dementia risk factors and cognitive decline in asymptomatic individuals in the early stage of preclinical AD and in individuals wit...

Descripción completa

Detalles Bibliográficos
Autor: Cumplido Mayoral, Irene
Tipo de recurso: tesis doctoral
Estado:Versión publicada
Fecha de publicación:2024
País:España
Institución:CBUC, CESCA
Repositorio:TDR. Tesis Doctorales en Red
OAI Identifier:oai:www.tdx.cat:10803/691460
Acceso en línea:http://hdl.handle.net/10803/691460
Access Level:acceso abierto
Palabra clave:Brain aging
Machine learning
Preclinical Alzheimer’s Disease
Neuroimaging
CSF biomarkers
Cognitive decline
Dementia risk factors
Pathophysiological mechanisms
Envejecimiento cerebral
Aprendizaje automático
Enfermedad de Alzheimer Preclínica
Neuroimagen
Biomarcadores de LCR
Deterioro cognitivo
Factores de riesgo de demencia
Mecanismos fisiopatológicos
Envelliment cerebral
Aprenentatge automàtic
Malaltia d'Alzheimer Preclínica
Neuroimatge
Biomarcadors de LCR
Deteriorament cognitiu
Factors de risc de demència
Mecanismes fisiopatològics
616.8
Descripción
Sumario:This thesis explores the links between neuroimaging-derived brain-age, which is computed using machine learning techniques, and different pathophysiological mechanisms, dementia risk factors and cognitive decline in asymptomatic individuals in the early stage of preclinical AD and in individuals with mild cognitive impairment. We showed that we can develop a neuroimaging-based biomarker for biological brain age (the so-called brain-age), which is robust and generalizable across participants from different cohorts. We further found that having an older-appearing brain is associated with higher neuronal loss measured with plasma neurofilament light (NfL), more advanced stages of amyloid and tau pathology and carrying the APOE-ε34 allele, and higher white matter hyperintensities. Additionally, brain-age might be able to show sex differences in brain aging. Moreover, findings showed that brain-age captures the association between modifiable risk factors and longitudinal cognitive decline, particularly in individuals that do not have Aβ pathology. Furthermore, our findings showed that higher TREM2-mediated microglial reactivity, as measured with CSF sTREM2, was associated with a younger brain-age after adjusting for AD pathology. These findings contribute to the growing field of brain-age as a biomarker of biological brain aging. Our results help in understanding of the mechanisms underlying biological brain aging, cognitive decline, and different physiological brain process such as neurodegeneration, glial activation, AD pathology and cerebrovascular disease. This highlights the potential of brain-age for preventive interventions targeting cognitive decline and providing insights into aging-related mechanisms.