An intuitive way to visualize brain response to bladder filling

Clarkson B1, Karim H1, Griffiths D J1, Resnick N1

Research Type

Basic Science / Translational

Abstract Category

Imaging

Abstract 290
Surgical Video 1
Scientific Podium Video Session 17
Thursday 30th August 2018
09:18 - 09:27
Hall D
Urgency Urinary Incontinence Urodynamics Techniques Basic Science
1. University of Pittsburgh
Presenter
B

Becky Clarkson

Links

Abstract

Introduction
The brain plays a vital role in the continence mechanism. Its involvement is complex, and our understanding can be confounded by the brain’s concurrent involvement in all other physiological processes. We have built our understanding of the brain’s role in bladder control upon a series of statistical analyses of activity during bladder-related tasks, yet these findings can be challenging to interpret and envisage, particularly for a non-neuroimaging audience. We sought to find a way to represent brain activity information that would intuitively represent a bladder task, such as filling, to a non-neuroimaging audience. Here we present an analysis of urodynamic bladder filling in an MRI scanner, from empty until strong desire to void, using a tool called ‘Regional Homogeneity’ (ReHo) [1] which assesses regional brain activation. ReHo evaluates the similarity of the activation pattern of each voxel with its nearest neighbors, essentially producing a map of activation over time which emphasizes robust clusters of activation. Since interpretation of brain imaging studies can be complex, we present our results as a time-lapse video of regional activation (co-occurrence of neighboring voxels) in the brain over the course of bladder filling to allow visualization of brain function and response to stimulus.
Design
Nine women over 60 years of age with urgency urinary incontinence (UUI) greater than five times per week underwent bladder filling at a constant rate of 50ml/minute until they signaled a strong desire to void, while concurrently having a BOLD fMRI scan on a 3T Siemens Trio MRI scanner. ReHo was then calculated for each voxel in the entire brain by calculating Kendall’s coefficient concordance (KCC) between neighboring voxels (neighbor was defined as any voxel that ‘touched’ a vertex – i.e., 27 voxel neighborhood), which measures the similarity between that voxel and its neighbors. Regions that are more concordant are more likely to have co-activated. We computed this metric over a sliding window of 20 data points – basically allowing us to understand ReHo as a function of time. We computed the average time series across all participants and then visualized the ReHo at the supplementary motor area (SMA), one of the regions of interest widely reported to be involved in the continence mechanism. We calculated ReHo at the last quartile of filling time and compared it to the initial three quartiles in each participant using a paired t-test to quantitatively assess activation changes.
Results
The video shows changes in ReHo occurring concurrently with a constant rate of bladder filling. The SMA activity increases substantially towards the end of filling. Quantitatively, we found that the mean ReHo in the SMA region of interest (MNI 2   0  48) was greater at the last 25% of the filling (mean=0.19, SD=0.01) compared to the first 75% of the filling (mean=0.17, SD=0.02) with t(9)=-2.44, p<0.05 (95% confidence interval [-0.028, -0.011]).
Conclusion
Time-lapse video provides a way to qualitatively assess brain function over time in response to bladder filling; this can provide some context to brain-bladder studies and a useful means of visualizing the brain’s role in continence. ReHo is a more complex measure than standard BOLD activation measures since it takes into account clusters of activation (i.e., neighboring voxels which activate together), and does not require a baseline measurement. This video shows that the SMA, known to be part of circuit 2 [2] and involved in urgency, acts as expected and can be visualized. We aim to use this method to both quantitatively assess activity and improve our ability to visualize and contextualize our data in future.
References
  1. Neuroscientist. 2016; 22(5): 486–505.
  2. Journal of Urology 2015; 194(3)708-15
Disclosures
Funding NIH R56 Clinical Trial No Subjects Human Ethics Committee University of Pittsburgh Institutional Review Board Helsinki Yes Informed Consent Yes