Storing urine and emptying the bladder is a multifaceted process that requires complex and coordinated activity as well as integration of peripheral afferent and central efferent signals at different levels. The human bladder has two main functions, namely, the storage of urine until an adequate capacity of the bladder is reached and micturition when an appropriate time and place is found to void. From a neural perspective, visceral sensory information is relayed from the bladder through the spinal cord and brainstem to higher brain areas, where levels of bladder fullness and urgency are monitored and our attention is shifted to our bladder when needed. Proper execution of these tasks is dependent on dedicated neuronal control systems at the level of the brain, brain stem and spinal cord.
	One of the dedicated nuclei at the level of the brain stem is the periaqueductal gray (PAG). The PAG is proposed to function as a relay station in micturition pathways in the central nervous system (CNS), and takes up a central location in the hierarchical system of lower urinary tract (LUT) control [1]. It is proposed that the PAG is responsible for relaying afferent information from the bladder to cortical and subcortical brain areas, while simultaneously acting as a gatekeeper relaying efferent information from cortical and subcortical areas to the pons and spinal cord [2]. 
	Optimal functioning of these central systems enables healthy adults to accurately assess their experienced levels of bladder fullness at any given time and to reliably predict for how long they will be able to postpone micturition. In conditions of LUT dysfunction, such as overactive bladder (OAB), participants might struggle with these tasks. The International Continence Society defines OAB as urgency, with or without urge incontinence, and often with frequency and nocturia. Over the past years, it has become more apparent that OAB is associated with altered bladder sensations. A disturbance in the neural processing of visceral sensory information might be related to this type of LUT dysfunction. Recent neuroimaging work has reported the first indications that PAG activity reflects subjective reported bladder sensations [3]. PAG activity might, therefore, offer insights into processing of visceral sensory information and could help identify alterations in interoceptive processes which might cause “false alarms” in patients with OAB.
	Animal research has indicated that the PAG is organized in a symmetrical columnar fashion, and the ventrolateral PAG and dorsolateral PAG are indicated to be involved in the control of voiding and storage of urine respectively. Previous research has indicated that the PAG can reliably be subdivided into distinct clusters of functional regions that can be differentiated based on resting state fMRI data at 7 Tesla. At the within-subject level, these clusters show a symmetrical organization and high level of similarity between empty and full bladder states [3], which is in line with what would be expected based on animal work.
	In the current study we have used 7 Tesla resting-state fMRI to assess the similarity between PAG organization at the group level and expected to observe a significantly higher spatial overlap between clusters from different participants than would be expected based on chance.

This study was designed and conducted in line with the Declaration of Helsinki and was approved by our local ethics committee. Written informed consent was obtained from each of our participants. For this follow up analysis we evaluated data from 10 female participants. After participants were positioned on the scanner bed we ran a T1-weighted whole brain anatomical scan using an MP2RAGE sequence. Next, we used a bladder filling protocol in which the bladder was filled with saline using a filling catheter (FR: 8) at a rate of 30 ml/min to evoke a strong desire to void in participants. We then ran a resting-state fMRI scan while participants experienced a strong desire to void in which we collected 420 T2*-weighted multiband echo planar imaging volumes (mb-EPI sequence, acceleration factor = 2, MB-factor = 2, TR = 1400ms, TE = 22ms, resolution = 1.1 x 1.1 x 1.1mm). We scanned 40 slices covering the supramedullary portion of the brain stem.
	Data were preprocessed using BrainVoyager and normalized to MNI space. PAG voxels were selected based on a mask covering the PAG of the MNI template. A voxel-by-voxel correlation matrix of the PAG was created and subdivided into clusters using the Louvain module detection algorithm. This algorithm outputs clusters with stronger within-cluster connectivity than between-cluster connectivity in blood-oxygen-level-dependent signal. The similarity of resulting clusters between participants was then assessed by computing the Dice similarity coefficient for all cluster comparisons across participants.
	Next, we ran 1000 permutations in which we created randomized correlation matrices based on the original data and subdivided these matrices, in the same approach as the original data, using the Louvain module detection algorithm. From these 1000 random organizational subdivisions we computed the Dice similarity coefficient for 100.000 randomly selected cluster comparisons. The Dice coefficients between cluster pairs were assessed statistically by ranking them to the Dice values observed in the permutations.

We observed a significantly higher similarity between cluster pairs across subjects compared to permutations. For 23 cluster combinations across participants we observed a higher similarity than could be expected based on chance (p = ≤0.05) (Fig. 1).

These results show that the PAG can be subdivided into distinct clusters which show a similar spatial distribution at group level. This analysis is a crucial step to determine the agreement between in vivo PAG organizational maps and the functional and anatomical organization of the PAG which is known from previous research. These advancements are necessary, first to be able to identify the relationship between LUT symptoms, such as urgency, and activity patterns in the PAG in normal and pathological situations, and determine interindividual conformity or diversity.
	Further investigation of how CNS activity patterns relate to subjective bladder fullness and urgency sensations can lead to identification of fMRI imaging biomarkers regarding OAB. A further unraveling of mechanisms of alarm falsification in OAB could potentially lead to new, non-invasive therapies like interoceptive bladder awareness training via bio-feedback. Moreover, this research may help us understand the underlying mechanisms of current therapies, such as sacral neuromodulation, and improve patient selection strategies.

The PAG is an important brain stem nucleus involved in sensation and control of the LUT and various other visceroceptive processes. Resting state fMRI data of the PAG can be used to subdivide this nucleus into clusters that show characteristics corresponding to anatomical studies of the PAG at the within-subject level. Here, we show that PAG clusters additionally show high spatial organizational similarity at the group level. This enables analysis of PAG activity related to bladder sensation and control at the group level as well as studies aiming to unravel the interaction between the PAG and the rest of the brain. Utilizing these approaches to study CNS changes in response to successful therapeutic interventions will not only help to improve current therapies and patient selection strategies, but also lead to the development of new therapies.

Fowler CJ, Griffiths D, de Groat WC. The neural control of micturition. Nat Rev Neurosci. 2008;9(6):453-466.
de Groat WC, Griffiths D, Yoshimura N. Neural control of the lower urinary tract. Compr Physiol. 2015;5(1):327-396.
de Rijk MM, van den Hurk J, Rahnama'i MS, van Koeveringe GA. Parcellation of human periaqueductal gray at 7-T fMRI in full and empty bladder state: The foundation to study dynamic connectivity changes related to lower urinary tract functioning. Neurourology and Urodynamics. 2021;40(2):616-623.