Connectomic analysis

Introduction
Connectomic analysis aims to understand on how brain connection inform functionality or brain disease pathophysiology. It view brain connections as a network and device network analysis approach to characterize its property.

Connectivity matrix and connectogram


source: 
Bernhardt, Boris C., SeokJun Hong, Andrea Bernasconi, and Neda Bernasconi. "Imaging structural and functional brain networks in temporal lobe epilepsy." (2013).


A connectivity can be derived by parcellating brain regions as "node" and tracking inter-regional connections as "edge". This creates a undirected graph of the brain network. 

The graph can be binary or weighted. A binary graph only concerns whether two regions are connected or not, whereas a weighted graph further considers connection strength.

A connectome generated from diffusion MRI fiber tracking is often called "structural" connectome.


Pros:
1. Study the brain structure as a graph.  
2. There are several readily available graph analysis approaches that can be used.

Cons:
1. Connectivity matrix is sensitive to tracking parameter and limited by the accuracy of the tracking algorithm.
2. How to partition the brain regions is a challenge


Graph theoratical analysis


Graph theoratical analysis quantifies the topology of a network as network measures, which allows for detecting aspects of functional integration and segregation, quantifying importance of individual brain regions, characterizing patterns of local anatomical circuitry, and testing resilience of networks to insult. 

Network measures can be categorized into global and local measures. Global measures quantifies property related to the entire graph, whereas local measures quantifies property of a node or an edge. Sometime, the mean of all nodal measures can be used as a global measure.

source: https://sites.google.com/site/bctnet/measures

A list of network measures in here.

Each dMRI scan gives a global network measure that can be compared between groups or correlated with any variables.

Reference
Rubinov, Mikail, and Olaf Sporns. "Complex network measures of brain connectivity: uses and interpretations." Neuroimage 52, no. 3 (2010): 1059-1069.
Bullmore, Ed, and Olaf Sporns. "Complex brain networks: graph theoretical analysis of structural and functional systems." Nature Reviews Neuroscience10.3 (2009): 186-198.


Local connectome fingerprint




"local connectome fingerprinting" (LCF) uses on q-space diffeomorphic reconstruction to calculate the density of diffusing water along major fiber bundles from diffusion MRI. 

LCF provides a vector measurement characterizing white matter structures at the voxel level for each dMRI scan. 

LCF was found to be highly unique for each individual, making it a tangible metric of connectomic differences. 

The variability in LCF was also found to be partially determined by genetic factors, but largely plastic across time. 

The density sampled exhibits high individuality that allows us reliably identify individuals or quantify similarity between two connectome architectures.



Pros 

1 LCF is highly specific to individuals
2 LCF may reveal neuroplasticity over time

Cons
1. High individuality brings high group variance
2. Analyzing LCF can be challenging

Reference
Yeh, Fang-Cheng, Jean M. Vettel, Aarti Singh, Barnabas Poczos, Scott T. Grafton, Kirk I. Erickson, Wen-Yih I. Tseng, and Timothy D. Verstynen. "Quantifying differences and similarities in whole-brain white matter architecture using local connectome fingerprints." PLoS computational biology 12, no. 11 (2016): e1005203.


Connectometry



Diffusion MRI connectometry uses permutation test to find out the association of white matter pathways with any study factor. It can be combined with a simple linear regression that includes all relevant variables (e.g. sex, site difference) in the model. The results will show tracks that exhibit positive or negative correlation with a variable of interest.


Connectometry adopts a "tracking the difference" paradigm, which is fundamentally different from the conventional paradigm of "finding the difference in tracks or regions". 

The analysis first identifies voxels that have strong correlations and tracks along axonal fiber directions to identify the consecutive fiber segment that shows continuous positive and negative correlation. The tracking is only limit to regions with high correlation, thus minimizing the noise contributed from other non-correlated regions.


Two types of connectometry analysis are available:

Group connectometry:

One is group connectometry [pdf], which identifies tracks associated with group difference or correlated with a study variable. It can also be used in a longitudinal study to study the difference between paired data (e.g. pre- and post- treatment). Group connectometry use a linear regression model and can include other variables to consider in the regression.

Individual connectometry:

Individual connectometry identifies damaged/enhanced pathways of an individual by comparing repeat scans of the same subject.

Alternatively, the subject's scan can be compared with a template to identifies connectivity that is below the average.

Pros:

1. Connectometry identifies only the affected segment of the fiber tracks
2. Allows for multiple regression, group comparison, paired group comparison.

Cons:
1. Cannot work on distorted brain structure (e.g. brain tumor)
2. Relatively poor at identifying short-ranged focal difference.

Reference:

Yeh, Fang-Cheng, David Badre, and Timothy Verstynen. "Connectometry: A statistical approach harnessing the analytical potential of the local connectome." NeuroImage 125 (2016): 162-171.
Yeh, Fang-Cheng, Pei-Fang Tang, and Wen-Yih Isaac Tseng. "Diffusion MRI connectometry automatically reveals affected fiber pathways in individuals with chronic stroke." Neuroimage: Clinical 2 (2013): 912-921.

Questions: 
1. What is the difference between QA and local connectome fingerprint?
2. Can connectometry analysis use FA and ADC?

Comments