dfConn (version 0.2.1)

lmmConn: Semiparametric modelling - Static

Description

Linear mixed effect modeling for static functional connectivity

Usage

lmmConn(dataList, op_lme)

Arguments

dataList

list, list of data matrices .

op_lme

list, options constructed by function options_lme, see options_lme.

Value

An object of list containing all the information of the static functional connectivity linear mixed model. It stores the information of model parameters, input data, and model results.

params

Store the parameters for the linear mixed model

parallel

Store the information about parallel computing environment and parameters

output_by_row

Store the model results information for data between each pair of regions

modelDyn_results

A combined dataframe with all the model results informations

est_CI

A list of matrices, contains information of confidence band estimate, see description below.

In each matrix of est_CI:

row 1

Base line estimate, difference between condition 1 and condition 2

row 2

Dynamic functional connecvity estimates for condition 1

row 3

Dynamic functional connecvity estimates for condition 2

row 4

Lower bound of 95% confidence interval of condition-difference estimates

row 5

Upper bound of 95% confidence interval of condition-difference estimates

row 6

Lower bound of 95% confidence interval of condition-1 estimates

row 7

Upper bound of 95% confidence interval of condition-1 estimates

row 8

Lower bound of 95% confidence interval of condition-2 estimates

row 9

Upper bound of 95% confidence interval of condition-2 estimates

Details

lmmDyn and lmmConn, we need to summarize the bootstrapping by taking the mean or median of the dFC estimate at each time point.

Examples

# NOT RUN {
# Assuming user has run MLPB_boot() and has summarized the bootstrapping results 
# by calculating the median or mean.

## Example with 5 subjects bootstrap-based static functional connectivity estimates data included
## Each subject's data has 731 time points in total, which includes 6 scans and 105 effective 
## time point for each scan. 
## 3 regions of interest (ROIs) comparision pairs are selected

data(MLPB_output_median)
subjects <- c('subject1', 'subject2', 'subject3', 'subject4', 'subject5')

# In our demo data, each subject has a scan with a total of 750 time points
time.points <- c(1:105, 126:230, 251:355,
                 376:480, 501:605, 626:730) 
                 
num.scan <- 6 # Each subject has 6 scans
ntps.per.scan <- 105 # Each scan has 105 time points

op <- options_lme(effective_tp = time.points, 
                 ntps.per.scan = ntps.per.scan,
                 subjects = subjects, 
                 num.scan = num.scan)
                 
resConn <- lmmConn(MLPB_output_median, op)
# }