Technical Name Characterisation of abnormal ageing based on resting state fMRI
Project Operator National Cheng Kung University
Project Host 吳馬丁
Summary
As the Taiwanese population ages, new methods are needed to decrease the emotional and healthcare cost of abnormal ageing, in particular dementia. We used feature selection and L1 regularized regression for identification of key relations among brain regions that can be attributed to abnormal ageing. Early detection of abnormal ageing is crucial for applying preventive therapies and hopefully stall or avoid the development of these diseases.
Scientific Breakthrough
Early detection of abnormal ageing is crucial for applying preventive therapies and hopefully stall or avoid the development of these diseases. We have established a model of normal ageing using L1 regularized regression and mapped the brains of 176 participants to a 264-region atlas and were able to identify key relations among these regions that can be attributed to either accelerated or delayed ageing.
Industrial Applicability
Dementia is estimated to have costed USD 818B worldwide in 2015 (1% of global GDP). Detection of abnormal ageing through non-invasive methods can lead to further tests to diagnosis dementia at its prodromal stages. Identifying the regions associated with delayed ageing can lead to development of technologies to enhance or improve the connectivity among these regions, effectively reducing age related cognitive deficits.
Keyword Resting state fMRI Age prediction LASSO Linear regression L1 regularization Default mode network Feature selection Parameter selection Variable selection Age effects
other people also saw