Human Brain Mapping: A predictive model based on resting state functional connectivity can predict neuropsychological test performance in later life

Neuropsychological tests are usually used to detect cognitive and functional changes in patients with dementia and related diseases.
It systematically describes the ability of subjects to perform specific cognitive tasks and assesses how subjects respond to activities in daily life
.

Neuropsychological tests also have a unique role in detecting and monitoring cognitive and functional changes related to dementia-related diseases

Recently, Human Brain Mapping magazine published research articles such as Jeanyung Chey, which explored the basis of brain function connections in neuropsychological tests in the elderly
.

The research aims to use predictive model methods to determine the best combination of functional connections and predict individual neuropsychological test scores
.

Use resting state functional connectivity and neuropsychological tests in the OASIS-3 data set (n = 428) to train the predictive model, and apply the identified model iterations to the internal test set (n = 216) and external test set (KSHAP, n = 151)

Descriptive statistics of OASIS-3 (internal verification set, n = 644) and KSHAP (external verification set, n = 151)

Descriptive statistics of OASIS-3 (internal verification set, n = 644) and KSHAP (external verification set, n = 151)

Descriptive statistics of neuropsychological test performance in the internal (OASIS-3) and external (KSHAP) validation data sets

Descriptive statistics of neuropsychological test performance in the internal (OASIS-3) and external (KSHAP) validation data sets

MRI scans were performed on all subjects to obtain resting functional magnetic resonance images, and 227 brain regions were selected for functional connection calculations
.

An overview of the schematic diagram of predictive modeling based on connectors
.

Use the training set of individual FC matrices (left) to identify the predicted weights of all connected features

An overview of the schematic diagram of predictive modeling based on connectors

Schematic overview of nested cross-validation
.

Internal verification (left): Triple cross-validation identifies the predicted weights adjusted by lambda in the sequence set

Schematic overview of nested cross-validation

The brain regions that mainly represent executive function and processing speed have higher prediction accuracy , such as fluency, TMT and digital symbol conversion, and previous studies have also shown that these tests have a higher value in predicting the process of dementia and the function of episodic memory
.

The prediction accuracy of the numerical breadth scale representing attention and working memory is poor .

And performing functions on behalf of the main brain region has a higher processing speed prediction accuracy representatives of working memory and attention span digital scale prediction poor accuracy vessel

The study shows that a predictive model based on resting state functional connectivity can predict the performance of neuropsychological tests in later life , but further evidence is needed to establish a theoretically effective clinical auxiliary diagnostic model

Resting state based prediction model can predict the function of connecting old age neuropsychological test performance prediction model resting state functional connection can be predicted based on the performance of old age neuropsychological test diagnostic

Original source

Kwak S, Kim H, Kim H, Youm Y, Chey J.

Kwak S, Kim H, Kim H, Youm Y, Chey J.
Distributed functional connectivity predicts neuropsychological test performance among older adults.
Hum Brain Mapp .
2021;42(10):3305-3325.
doi:10.
1002/hbm.
25436 Kwak S , Kim H, Kim H, Youm the Y, Chey J.
Distributed Functional Connectivity Predicts Performance Test neuropsychological Among Adults older.
  Hum Brain Mapp in this message