The　Ε4　allele　of　the　apolipoprotein　E　(APOE)　gene　is　well　established　as　a　strong　genetic　risk　factor　for　the　development　of　late-onset　Alzheimer’s　disease　(AD)　and　declining　cognitive　function,　with　AD　patients　showing　a　higher　frequency　of　one　or　two　copies　of　the　Ε4　allele　than　in　the　healthy　elderly　population.　The　impact　of　this　allele　on　cognitively　healthy　APOE-Ε4　carriers　a　brain　pattern　that　resembles　the　pattern　seen　in　AD　patients　has　been　documented,　but　its　influence　on　normal　brain-aging　patterns　is　still　elusive.　Capitalizing　on　a　cross-sectional　study　on　113　cognitively　healthy　late　middle　aged　and　older　adults　with　two,　one　and　no　APOE-e4　alleles,　the　current　study　aims　to　characterize　the　ability　of　comparing　a　person＇s　estimated　brain　age　with　their　chronological　age　to　identify　the　potential　effects　of　the　APOE-Ε4　on　brain-aging　morphology.　The　brain　age　estimation　model　was　trained　on　magnetic　resonance　image　(MRI)　data　from　594　cognitively　normal　late　middle　aged　and　older　individuals,　by　using　the　well-known　AlexNet,　with　relevance　vector　regression　methods.　The　model　achieved　R2=　0.79,　and　mean　absolute　error　of　4.51　years.　When　applied　the　pretrained　prediction　model　to　new　subjects,　the　estimated　brain　age　of　subjects　in　Ε4　homozygotes　was　2.37　±　3.65　years　greater　than　their　chronological　age　(p=　0.002),　whereas　within　the　control　group,　estimated　brain　age　was　similar　to　the　chronological　age.　The　result　showed　that　healthy　APOE-Ε4　homozygotes　have　accelerated　brain-aging　pattern,　which　may　suggest　dose-dependent　disease　vulnerability　on　the　brain　structure　level.　The　proposed　method　can　be　used　as　a　potential　imaging　biomarker　for　detecting　the　earliest　effects　of　AD　on　the　brain　of　cognitively　unimpaired　people.　©　2019　Association　for　Computing　Machinery.