|
KMID : 0385620200240020001
|
|
Korean Journal of Psychopathology
2020 Volume.24 No. 2 p.1 ~ p.8
|
|
|
Brain Age, New Neuropathologic Marker
|
|
Kim Myung-Ju
Hwang Jae-Uk
Woo Sung-Il
|
|
|
Abstract
|
|
|
|
The human brain undergoes aging throughout one’s lifetime after its maturation period. However, the patterns of brain aging vary from person to person. Recently, the brain age prediction brings our attention increasingly since it may be useful as a biological marker of the level of brain aging. Estimation of brain age is of great clinical significances, since it can suggest the severity of neurological diseases as well as psychiatric disorders. It has been reported that regional volumes of cerebral structures and cortical thickness would be decreased with aging. Yet, these changes vary depending on numerous factors such as tissue and area of the brain aging and accelerate after a certain period of time. In order to establish brain age prediction algorithms with high accuracy, it is necessary to select relevant input data and algorithms. Majority of the previous studies trained the models through volume-based data and in particular, the study applying the latest prediction algorithms, such as convolutional neural networks, has shown reasonably high accuracy with a mean absolute error of around 4 years. Growing body of literatures regarding the brain age prediction has been available since 2015. In the future, more accurate model will be developed with more advanced machine learning and larger number of brain imaging data. If the model would be developed for younger population, its clinical significance will be increased.
|
|
|
KEYWORD
|
|
|
brain age, MRI, machine learning
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|
|