Comprehensive Non-invasive Analysis of Lower Urinary Tract Anatomy Using MRI

Abstract

Purpose

Anatomic changes that coincide with aging including benign prostatic hyperplasia (BPH) and lower urinary tract symptoms (LUTS) negatively impact quality of life. Use of MRI, with its exquisite soft tissue contrast, full field-of-view capabilities and lack of radiation is uniquely suited for quantifying specific lower urinary tract features and providing comprehensive measurements such as total bladder wall volume (BWV), bladder wall thickness (BWT) and prostate volume (PV). We present a technique for generating 3D anatomical renderings from MRI to perform quantitative analysis of lower urinary tract anatomy.

Methods

T2-weighted fast-spin echo MRI of the pelvis in 117 subjects (59F;58M) aged 30–69 (49.5±11.3) without known lower urinary tract symptoms were retrospectively segmented using Materialise software. Virtual 3D models were used to measure BWV, BWT and PV.

Results

BWV increased significantly between the 30–39 and 60–69 year age group in women (p=0.01), but not men (p=0.32). BWV was higher in men than women ages 30–39 and 40–49 (p=0.02, 0.05, respectively) but not 50–59 or 60–69 (p=0.18, 0.16, respectively). BWT was thicker in men than women across all age groups. Regional differences in BWT were observed both between men and women and between opposing bladder wall halves (anterior/posterior, dome/base, left/right) within each sex in the 50–59 and 60–69 year groups. PV increased from the 30–39 to 60–69 year groups (p=0.05). BWT was higher in subjects with enlarged prostates (>40cm³)(p=0.05).

Conclusion

Virtual 3D MRI models of the lower urinary tract reliably quantify sex-specific and age-associated changes of the bladder wall and prostate.

Introduction

Prior studies have shown that lower urinary tract symptoms (LUTS) in aging men and women are associated with anatomical changes in the lower urinary tract[1], [2]. These changes include thickening of the bladder wall and prostate enlargement secondary to benign prostatic hyperplasia (BPH)[3]–[5]. LUTS is a significant societal problem that substantially impacts quality of life[4]. Symptoms commonly include urgency, frequency, retention, nocturia, hesitancy and incomplete emptying. Increased bladder wall thickness (BWT) in association with overactive bladder in women and BPH/LUTS in men has been previously demonstrated with ultrasound (US) and computed tomography (CT)[1], [2], [12], [13], [3], [4], [6]–[11]. In these studies, parameters such as BWT and bladder volume were quantified by localized measurements on a 2D plane[2], [3], [6]–[13] and mathematical estimation[6], [9]. US, however, is limited by its lack of full field-of-view capabilities and CT has poor tissue contrast compared to MRI and exposes the patient to radiation. Further, neither of these imaging modalities can provide reliable and reproducible quantitative measurements of the lower urinary tract.

MRI, as a potential solution, offers several advantages, including exquisite soft tissue contrast, full-field of view capabilities and improved patient safety profile relative to CT. Superior soft tissue contrast allows for quantitative characterization of the bladder wall and prostate zonal anatomy. Given that it is unknown whether bladder wall thickness is homogeneous throughout the organ, a full field of view would allow for measurement of potential regional variation. Further, no exposure to radiation reduces risks to the patient. However, the use of MRI to examine lower urinary tract anatomy is limited. We present a quantitative volumetric methodology for assessing the anatomy of the lower urinary tract and report the normative values for bladder wall volume (BWV), BWT and prostate volume (PV) in men and women without LUTS. The purpose of this study is to provide a method to correlate lower urinary tract anatomic structure and function (and dysfunction) in men and women with LUTS. This method also provides a first-ever comprehensive survey of changes in lower urinary tract anatomy with aging and thereby establishes reference standards.

Materials and Methods

Patient Population

117 pelvic MRIs (58 male) of subjects age 30–69 [mean age of 50.2±11.4 years (men), 48.9±11.3 (women) and 49.5±11.3 years (entire population)] were retrospectively analyzed after IRB approval and following a HIPAA compliant protocol. Subjects were grouped by age: 30–39, 40–49, 50–59 and 60–69 for both men (n_30–39=14, n_40–49=15, n_50–59=15, n_60–69=14) and women (n_30–39=16, n_40–49=14, n_50–59=15, n_60–69=14).

MRI examinations were performed on patients at the University of Wisconsin (UW) Health system between the dates of June 1, 2017 and May 31, 2018 for indications unrelated to the lower urinary tract such as gynecologic malignancy, cancer or hip or back pain. Exclusion criteria included: below 30 years of age, above 69 years of age or an MRI without complete coverage of the bladder and prostate.

Fast-spin echo (FSE) T2-weighted axial acquisitions with the following parameters were utilized for analysis: TR=4119ms; TE=92.4ms; matrix=320×256; FOV=260×234mm²; slice thickness=3mm; true spatial resolution=0.81×1.02mm; flip angle=90°; receiver bandwidth=122.07±31.25kHz; field strength=1.5 or 3T.

MRI Analysis

Axial FSE T2-weighted acquisitions, in DICOM (Digital Imaging and Communications in Medicine) format, were uploaded into version 19.0 of the design software “Mimics” (Materialize, Leuven, Belgium, 2016), used to extract virtual anatomic 3D models from medical images. The bladder lumen, bladder wall and whole prostate (men) were individually segmented in the axial plane by an author in consensus with an experienced engineer and abdominal radiologist. Figure 1

Fig 1:

Segmentation of the bladder and prostate was performed with Mimics software. Select contoured images of the bladder (a), bladder and prostate (b) and prostate (c).

After contouring the relevant organ on each MRI slice, a mask of the segmented anatomy was constructed. The mask was then converted into a 3D model using the “Calculate 3D from Mask” tool. This process was repeated separately for the bladder lumen and wall together (3D entire bladder), a second time for the bladder lumen (3D lumen model) and a third time for the whole prostate which provided the prostate volume (PV). The bladder urine volume was recorded from the 3D lumen model. Then, the bladder wall was then rendered as a 3D model by subtracting the 3D lumen model from the 3D entire bladder model providing bladder wall volume (BWV). Figure 2 Bladder wall thickness (BWT) was measured manually (serosa to mucosa) at the anterior (BWT_Ant) bladder wall, similar to previous studies[2], [8]–[13] and at the site of maximum thickness (BWT_Max) on the axial acquisition. Both BWT_Ant and BWT_Max, were measured three times and averaged.

Fig 2:

The entire bladder (lumen and wall) was contoured (a). Then, the bladder lumen was contoured (b). The bladder wall (c) was rendered by subtracting the bladder lumen (b) from the entire bladder (b) mask.

Bladder Wall Thickness Quantification

3D models of the bladder wall were exported from Mimics into 3-matic, version 11.0 (Materialize, Leuven, Belgium, 2016) as STL files for manipulation and quantification of the 3D virtual model. A thickness gradient tool was used on the 3D models of the bladder wall. The computational mean and median BWT of the entire bladder were collected for each subject. Averages of the mean and median values were calculated for men and women in each age group.

The 3D model of the bladder with its center point, calculated in Mimics, were imported together into 3-Matic. There, the bladder was divided along the center point into six opposing halves: anterior and posterior, dome and base, and left and right sidewalls. Figure 3 Regional differences in mean and median BWT were recorded.

Fig 3:

Bladder wall before (left) and after (right) segmentation into anterior/posterior, dome/base and left/right divisions performed using 3-matic software. Gradient shows thinnest wall in green, medium wall thickness in yellow and thickest wall in red.

Each parameter was summarized by median (interquartile range). Differences between male and female subjects were analyzed using t-tests or ANOVA F-tests. Significance level was set at 0.05.

Male bladder data was stratified into two groups based upon prostate volume, above (n=12) and below (n=46) 40cm³, a cutoff previously used to determine risk for BPH[5]. Comparisons of BWV and BWT were made between the two groups.

Results

Bladder Wall Volume (BWV)

BWV was greater in men than in women in age groups 30–39 and 40–49 (p=0.02 and 0.05, respectively) but not age groups 50–59 and 60–69 (p=0.18 and 0.16, respectively). Table 1

Table 1:

Measurements for male and female bladder wall volume (cm³) stratified by age.

BWV	Men		Women		P-value
Age	Median (cm3)	IQR (cm3)	Median (cm3)	IQR (cm3)
30–39	33.4	27.8–45.8	29.0	26.9–31.1	*0.02
40–49	36.8	32.2–49.8	28.9	23.2–37.2	*0.05
50–59	38.4	35.7–49.1	30.8	22.5–43.9	0.18
60–69	37.0	32.7–50.5	35.3	27.1–42.8	0.16

Median (IQR) BWV did not increase significantly with age from the 30–39 to 60–69 year groups in men [33.4 (27.8–45.8) to 37.0 (32.7–50.5) cm³, p=0.32]. Median (IQR) BWV increased significantly with age from the 30–39 to 60–69 year groups in women [29.0 (26.9–31.1) to 35.3 (27.1–42.8) cm³, p=0.01].

Bladder Wall Thickness (BWT)

The urine volume (bladder distention) was similar in men and women across the age groups (84.1±59.6 vs 74.1±68.5cm³, p=0.40) even though we were unable to control for urine volume in the retrospective design. Patients at our institution are asked to void prior to MRI scanning which likely accounts for this finding. When manually measured, BWT_Ant in men and women was not significantly different for any age group. BWT_Max was significantly higher in men for the 50–59 and 60–69 year groups (p=0.01 and 0.005, respectively) but not in the 30–39 and 40–49 groups. Table 2

Table 2:

Male and female anterior and maximum bladder wall thickness (mm) stratified by age when measured manually.

	Median (IQR) Anterior BWl (mm)			Median (IQR) Maximum BWT (mm)
Age	Men	Women	P-value	Men	Women	P-value
30–39	3.9 (3.1–5.1)	3.5 (2.7–4.2)	0.36	5.1 (4.2–6.4)	4.5 (3.8–5.7)	0.31
40–49	3.5 (3.2–4.3)	4.5 (3.6–4.8)	0.09	5.5 (4.5–6.4)	5.2 (4.7–5.8)	0.48
50–59	4.2 (4.0–5.0)	3.4 (2.6–4.9)	0.07	6.4 (5.3–7.2)	5.1 (4.1–5.9)	0.01
60–69	5.3 (4.5–6.3)	3.8 (3.1–5.1)	0.06	6.9 (6.6–8.1)	4.7 (3.8–6.3)	<0.01

Both BWT_Ant and BWT_Max, increased significantly with age from 30–39 to 60–69 year groups in men (p=0.006 and 0.002, respectively) but not in women (p=0.26 and 0.42, respectively).

With automated measurements, median BWT generated from the 3D bladder model was significantly higher in men than women for the 50–59 and 60–69 year groups (p=0.01 and 0.03, respectively) but not the 30–39 and 40–49 year age groups (p=0.35 and 0.24, respectively). Median male BWT was higher in every measured region (base, dome, right, left, anterior and posterior) for every age group compared to women. This difference was statistically significant for the majority, but not all, of the bladder regions for the age 50–59 (p=0.001, 0.03, 0.06, 0.0006, 0.02 and 0.03, respectively) and continued to be significant for the base, left and anterior bladder for the 60–69 age group (p=0.0009, 0.02, and 0.03, respectively). Tables 3a–c

Table 3a:

Regional median (IQR) bladder wall thickness (mm) for males stratified by age.

Age	Base	Dome	Right	Left	Anterior	Posterior
30–39	2.8 (2.1–3.2)	2.8 (2.0–3.3)	2.6 (1.8–3.2)	3.1 (2.3–3.6)	3.1 (2.4–3.5)	2.6 (2.0–3.2)
40–49	2.7 (2.1–3.2)	3.0 (2.5–3.7)	2.8 (2.1–3.2)	3.0 (2.4–3.5)	3.1 (2.5–3.7)	2.6 (2.1–3.2)
50–59	3.3 (2.9–3.8)	2.4 (2.2–2.8)	2.8 (2.4–3.1)	3.3 (2.9–3.7)	3.4 (2.9–3.9)	2.6 (2.2–2.8)
60–69	3.7 (3.2–4.3)	2.6 (2.3–3.1)	3.0 (2.3–3.6)	3.5 (2.8–4.4)	3.9 (3.0–4.4)	2.7 (2.1–3.2)

Table 3c:

Test of significant difference of wall thickness between opposing bladder halves for males and females stratified by age.

P-values between male and female populations
	Base	Dome	Right	Left	Anterior	Posterior
30–39	0.41	0.14	0.77	0.11	0.48	0.19
40–49	0.50	0.05	0.43	0.11	0.37	0.15
50–59	<0.01	0.03	0.06	<0.01	0.02	0.03
60–69	<0.01	0.47	0.08	0.02	0.03	0.06

In men, BWT of opposing bladder halves (dome/base, right/left and anterior/posterior) differed in the 50–59 year age group (p=0.001, 0.04 and 0.003, respectively) and dome/base and anterior/posterior remained significant in the 60–69 age group (p=0.003 and 0.001, respectively). In women, BWT of opposing bladder halves differed between dome/base and anterior/posterior bladder for the 50–59 age group (p=0.04 and 0.07, respectively) and anterior/posterior in the 60–69 cohort (p=0.04).

Prostate Volume (PV)

Mean prostate volume was 28.7±7.8cm³ for men in the 30–39 year age group, which increased to 47.0±31.7cm³ in the 60–69 year age group (p=0.05).

Male data was also stratified into two groups based on PV: <40cm³ (n=46) and >40cm³ (n=12). The mean ages for the two groups were 48.2±11.3 and 57.9±8.3 years, respectively. Median BWV (IQR) was 36.6cm³ (30.1–47.8) in the <40cm³ group and 42.6cm³ (37.4–51.5) in the >40cm³ group (p=0.40). Median BWT generated from the 3D bladder model was significantly thinner in the <40cm³ group compared to the >40cm³ group [2.90cm³ (2.2–3.5) vs 3.46cm³ (3.1–4.0), p=0.05]. Further, there were regional differences between men with and without enlarged prostates. BWT was higher in the >40cm³ group in the base, left and anterior bladder halves (p=0.01, 0.04 and 0.03, respectively). Figure 4

Fig 4:

Comparison of bladder wall thickness (mm) of the whole and segmented bladder stratified by prostate volume. P-values for each group, from left to right, are 0.05, 0.01, 0.58, 0.11, 0.04, 0.03 and 0.30.

Discussion

We implemented a new quantitative method of measuring anatomic parameters of the lower urinary tract in men and women with MRI. We contoured relevant organs of the lower urinary tract and created a 3D virtual model, from which we measured the bladder wall volume, bladder wall thickness and prostate volume. We found that BWV in men <50 was higher in men than women. This is consistent with Oelke et al., who found detrusor wall thickness in healthy men to be significantly greater than in healthy women using US[12]. However, BWV increased with aging at a greater rate in women than in men. As a result, BWV in men and women >50 was similar. Anterior and maximum BWT increased with age and was higher in men than women >50. Importantly, the bladder base and not the anterior bladder wall thickness increased the most with age. We found regional variation in BWT for both men and women. However, regional variation in BWT is substantially higher in men and worsens with aging. As expected, we found that PV increased in aging men. It is well known that PV directly correlates with age and with LUTS. Williams et al. reported an average prostate volume of 32.9±2.9cm³ in men aged 30–45 using MRI (n=9)[14], similar to the mean prostate volume of 28.7±7.80cm³ in our group of men aged 30–39. In Franco et al.[7], an older cohort of men with a mean age of 67±8.2 years had a PV of 53±33cm³. Our group of men aged 60–69 had a similar mean PV of 47.0±31.7cm³. Lastly, we found that BWT but not BWV was greater in men with enlarged prostates.

Our results are consistent with previous findings for the various age groups. Hakenberg et al, using US, reported a mean BWT of 3.25±0.13mm in a group of 75 men without LUTS aged 21–40[9] compared to a mean BWT of 2.9±1.1mm in our 30–39 year male group. Kuo reported detrusor wall thickness in an empty female bladder (median age 46.7±12.5 years) to be 4.73±0.97mm along the anterior wall and 3.83±1.06mm posteriorly[15] compared to 2.8±1.1mm and 2.2±0.8mm in our study group of women aged 40–49, respectively. Importantly, Kuo’s BWT values decreased with increasing bladder volume suggesting the difference in BWT between studies is likely due to under-distention of the bladder. Ucer et al. reported a BWT of 2.01±.06mm for a full bladder and 3.25±1.01mm for an empty bladder for a cohort of 31 healthy women aged 44.21±11.60 years[16].

Our work demonstrates that MRI can be used to quantitatively assess clinically relevant anatomic parameters of the lower urinary tract, with a new focus on regional variations in the bladder wall. Previous studies have used MRI to investigate bladder shape[17] and BWT using normalized thickness models[18]. We successfully implemented a reproducible technique to measure and compare parameters including BWV, BWT, and PV. Importantly, these parameters can be correlated with symptomatic conditions such as over-active bladder[19], under-active bladder[20] and bladder outlet obstruction[12], [13], [21]. This methodology has several potential advantages. Relative to current standard of care imaging of the lower urinary tract, MRI offers improved soft tissue contrast without radiation exposure and full-field of view, volumetric imaging of the bladder and prostate. Therefore, MRI has a potential for use in longitudinal patient studies. Further, creation of volumetric models may allow for improved quantification. For example, the ability to manipulate a 3D model allowed us to investigate regional differences in BWT, which may be of clinical value in patients with LUTS. Multichannel urodynamics, the standard method for clinical evaluation of LUTS offers a functional test of the bladder. The quantitative MRI methods introduced in this study complements the functional information provided by urodynamics with anatomical information allowing a correlation of bladder structure and function.

Manual segmentation and measurement of BWT and PV are labor intensive, require anatomic expertise and are subject to variability. Importantly, we minimized the risk of inconsistent segmentation by limiting measurements to a single investigator, which may not be possible in the clinical setting. Previous studies have used automated, formula-based methods for collection of volume parameters. Automated methods of segmenting the anatomy, once developed, may save time and improve measurement consistency. Disadvantages of automated segmentation include variability between different computational segmenting programs. Formula based segmentations may be subject to false assumptions (especially with irregular anatomy) and the need to manually review the segmentation to ensure accurate readings[22]–[24].

Our study has several limitations. The retrospective design did not allow for MRI pulse sequence optimization. While the ideal sequence for bladder wall and prostate segmentation is unknown, we have found, in subsequent work, that an isotropic T2-weighted FSE without fat saturation provides improved bladder wall and prostate contrast on both acquired and reformatted images. While our sample size was adequate to support proof of concept, it is insufficient to detect small differences in BWT, BWV and PV. Further, we were not able to control for bladder distention due to filling status in this retrospective study, a variable shown to have direct correlation with BWT[12]. However, there was no difference in post void residual between men and women in our study.

Conclusion

Pelvic MRI with segmentation and 3D anatomical renderings provide a novel method for quantifying important parameters of the lower urinary tract including bladder wall thickness, bladder wall volume and prostate volume.

Table 3b:

Regional median (IQR) bladder wall thickness (mm) for females stratified by age.

Age	Base	Dome	Right	Left	Anterior	Posterior
30–39	2.6 (2.1–2.8)	2.3 (1.9–2.4)	2.5 (2.1–2.8)	2.5 (1.9–2.7)	2.8 (2.2–3.0)	2.2 (1.7–2.4)
40–49	2.5 (1.8–2.8)	2.3 (1.7–2.7)	2.5 (1.9–2.8)	2.5 (1.7–2.7)	2.8 (2.1–3.3)	2.2 (1.6–2.9)
50–59	2.4 (2.0–2.7)	1.9 (1.6–2.1)	2.3 (1.7–2.4)	2.4 (2.0–2.8)	2.6 (2.1–2.9)	2.1 (1.6–2.5)
60–69	2.5 (2.0–2.8)	2.4 (1.6–3.2)	2.5 (1.8–3.0)	2.7 (1.8–3.6)	2.9 (2.2–3.9)	2.2 (1.6–2.6)