Wearable Ultrasound Devices for Continuous Bladder Monitoring: Emerging Utility and Current Limitations in Neurourology

Abstract

Monitoring bladder filling is crucial in patients with neurogenic lower urinary tract dysfunction, particularly for safe and timely clean intermittent catheterization. Wearable ultrasound devices have recently gained attention as potential aids for continuous bladder monitoring, but they differ fundamentally from traditional bladder scanners. While portable scanners directly measure postvoid residual volume in milliliters, first-generation wearable devices using amplitude mode (A-mode) ultrasound primarily monitor bladder fullness trends and provide threshold-based alerts rather than precise quantitative volume measurements. This narrative review examines wearable ultrasound technologies in neurourology, searching PubMed, Embase, IEEE Xplore, and Google Scholar for relevant clinical studies. Most available data come from small observational studies in pediatric populations. In children with functional voiding dysfunction (FVD), devices such as SENS-U achieved sensitivities of 85%–90% for detecting bladder fullness thresholds with good correlation to urodynamic measurements. However, performance in neurogenic bladder populations is concerning: a recent study in children with neurogenic bladder reported only 46% accuracy and 13% specificity using DFree, indicating unacceptably high false-positive rates. No prospective validation studies have been conducted in adults with spinal cord injury, multiple sclerosis, or those at risk for autonomic dysreflexia. Regulatory classifications vary significantly among devices: SENS-U holds Conformité Européenne marking as a Class IIa medical device for pediatric use, while DFree is marketed as a consumer wellness product without medical device approval. EdgeFlow UW20, which employs a deep learning-based automatic bladder contour detection, recently received Ministry of Food and Drug Safety approval in Korea as a digital medical device. Currently, evidence supports considering wearable bladder monitors only for pediatric FVD under medical supervision. For neurogenic bladder populations, these devices are not recommended for routine clinical use until population-specific validation studies demonstrate adequate diagnostic accuracy. The emergence of B-mode (brightness mode) deep learning-based devices may address the fundamental limitations of first-generation A-mode technology.

INTRODUCTION

Lower urinary tract dysfunction affects many people with neurological disorders. About 70%–84% of those with spinal cord injuries (SCIs) develop bladder problems, and in patients with multiple sclerosis (MS), the rate can be even higher—up to 90% over the course of their disease [1-3]. If bladder management is neglected, complications can arise quickly: recurrent urinary tract infections (UTIs), upper urinary tract damage, and in SCI patients, potentially life-threatening episodes like autonomic dysreflexia (AD) [4, 5].

Clean intermittent catheterization (CIC) remains the standard approach for managing neurogenic bladder [5, 6]. However, determining the optimal timing for catheterization is challenging. Portable bladder scanners, such as the BVI 3000 or BVI 9400 (Verathon Inc., Bothell, WA, USA), assist clinicians in estimating postvoid residual (PVR) urine volume. These devices typically apply fixed clinical thresholds, often 100 or 200 mL, and validation studies have demonstrated strong correlations between scanner measurements and catheterized bladder volumes, with reported correlation coefficients ranging from r=0.79 to 0.95 [7, 8]. Nevertheless, bladder scanners have limitations— they require skilled operators, provide only singlepoint measurements, and are not designed for continuous home monitoring.

Wearable ultrasound devices represent a new development in this area. Most first-generation wearable devices do not report an exact PVR value in milliliters; instead, they continuously monitor bladder fullness trends and can alert users when their bladder reaches a certain percentage—typically around 70%– 80% of estimated capacity. This represents a fundamental shift in the conceptual approach to bladder monitoring. Wearables are not meant to replace traditional measurements, but rather to supplement them—to serve as an additional tool, not a replacement [9-12]. This review examines the current evidence for wearable bladder monitoring devices in neurourology.

METHODS

A comprehensive narrative review was conducted to evaluate the current clinical readiness of wearable ultrasound devices for bladder volume monitoring [13]. Literature searches were performed using PubMed, Embase, IEEE Xplore, and Google Scholar for articles published between January 2010 and February 2026.

Search terms included combinations of “wearable bladder monitor,” “bladder volume,” “ultrasound,” “urinary retention,” “post-void residual,” and “continuous monitoring.” Studies were eligible if they (1) involved human subjects, (2) evaluated wearable or body-mounted bladder monitoring devices, and (3) reported diagnostic accuracy, agreement with reference standards, or clinically relevant performance metrics.

Studies focusing exclusively on conventional handheld bladder scanners without wearable components, animal-only experiments, or simulation-only studies were excluded. When available, catheterized bladder volume measurement was considered the reference standard because it represents the most direct and widely accepted method for determining actual bladder volume in clinical validation studies. Although bladder fullness or functional bladder capacity may vary among individuals and may also be assessed using urodynamic findings or clinical symptoms, catheterization provides an objective measurement of true bladder volume and is therefore commonly used as the reference standard in validation studies of bladder monitoring devices. Given the heterogeneity of study designs and outcome measures, a formal meta-analysis was not performed.

For clarity, the strength of evidence for each device was categorized based on study design and validation rigor, ranging from exploratory feasibility studies to prospective clinical validation against catheterized volumes. These classifications are summarized in Tables 1–3.

Table 1.

Summary of wearable ultrasound-based bladder monitoring devices and validation studies

Device	Study design	Population	Reference standard	Outcome measure	Key finding
SENS-U [14]	Diagnostic accuracy	Pediatric FVD (n=30)	Urodynamic infusion	Sensitivity for fullness	90%, rs=0.94
SENS-U [15]	Clinical feasibility	Pediatric FVD (n=68)	Clinical assessment	Notification accuracy	91% accurate
DFree [20]	Diagnostic accuracy	Pediatric NGB (n=40)	CMG volume	Accuracy, Sens, Spec	46%, 70%, 13%
EdgeFlow UW20 [22]	Ongoing clinical trial	Postoperative (planned)	Catheterization	Pending publication	Data not yet available
Bladder Scan a) [7, 8]	Validation studies (selected)	Mixed (selected populations)	Catheterization	Correlation	r=0.79–0.95

Evidence levels are categorized according to study design and reference standard.

FVD, functional voiding dysfunction; NGB, neurogenic bladder; CMG, cystometrogram.

^a)Conventional portable scanner included for reference.

Table 2.

Evidence classification and clinical requirements for wearable bladder monitoring

Clinical population	Evidence level	Device output type	Minimum clinical requirement	Clinical recommendation
Pediatric functional voiding dysfunction	Moderate	Fullness alert	Sensitivity ≥85%	May consider as adjunct to urotherapy
Pediatric neurogenic bladder	Low (negative)	Fullness alert	Specificity ≥80% required	Not recommended
Adult SCI/MS	No validation data	Quantitative volume	Sensitivity ≥95% for overdistension detection	Validation required
Autonomic dysreflexia risk	No validation data	Quantitative volume	High sensitivity critical	Contraindicated pending validation
Neobladder/diversion	No evidence	Quantitative imaging	Anatomical reliability uncertain	Not applicable

Evidence classifications are based on currently available literature and expert interpretation, as prospective validation studies remain limited.

These classifications are based on currently available studies and expert interpretation, as prospective validation studies for wearable bladder monitoring devices remain extremely limited.

SCI, spinal cord injury; MS, multiple sclerosis.

Table 3.

Commercial devices: regulatory status and intended use classification

Device	Manufacturer	Regulatory status	Intended use classification	Availability
SENS-U	Novioscan (Netherlands)	CE Mark Class IIa (March 2018)	Medical device for pediatric use	Europe only
DFree	Triple W (Japan/USA)	No FDA clearance; marketed as consumer product	Consumer wellness device a)	USA, Japan, Europe
EdgeFlow UW20	Edgecare (Korea)	MFDS digital medical device (April 2025)	Medical device for monitoring	Korea (trials ongoing)

FDA, U.S. Food and Drug Administration; MFDS, Ministry of Food and Drug Safety.

^a)Manufacturer explicitly states: “DFree is not a medical device.” Adapted from DFree website [21].

TECHNICAL PRINCIPLES AND TERMINOLOGY

Before reviewing clinical evidence, it is useful to clarify several terms used throughout this review, including “wearable bladder scanners,” “continuous bladder monitoring,” and “alert-based wearable devices.” PVR measurement refers to quantifying the amount of urine remaining in the bladder after voiding, typically measured in milliliters with established clinical thresholds. In this review, the term ‘wearable bladder scanner’ refers to body-mounted ultrasound devices intended for continuous bladder monitoring. Most currently available systems provide bladder fullness trends or threshold-based alerts, whereas emerging imaging- or array-based platforms aim to estimate quantitative bladder volume. In contrast, bladder fullness monitoring describes how full the bladder is, either continuously or at intervals, and is usually reported as a percentage of estimated bladder capacity or in broad categories. This approach is primarily used to guide the timing of voiding or catheterization, and most currently available wearable devices focus on this function.

Commercial wearable devices utilize amplitude mode (Amode) ultrasound, which employs single or multiple transducers. This differs from the brightness mode (B-mode) imaging used in conventional ultrasound scanners. A-mode generates a 1-dimensional (1D) trace of echo amplitude versus depth (time-of-flight) along a single ultrasound beam (scan line), or along several predefined fixed beam lines when multiple transducers are used. It can detect reflections from the bladder walls, but it does not inherently reconstruct a 2-dimensional (2D) image; therefore, bladder volume is typically inferred from geometric assumptions and proprietary algorithms rather than measured directly. Therefore, most wearable A mode devices report relative “fullness” (e.g., levels or % capacity) rather than an absolute volume in milliliters. In contrast, B-mode scanners generate 2D images and use geometric calculations to estimate bladder volume with established accuracy. The fundamental differences between A-mode and B-mode ultrasound principles used in bladder monitoring are illustrated in Fig. 1.

Fig. 1.

Schematic illustration of A-mode and B-mode ultrasound for bladder monitoring. (A) A-mode: echo amplitudes from bladder walls plotted as spikes; wall-to-wall echo distance represents bladder diameter. (B) B-mode: multiline scan converts echo amplitudes into gray-scale brightness dots, enabling 2-dimensional bladder diameter and volume estimation.

The SENS-U (Novioscan, The Netherlands) device uses four piezoelectric transducers with a 30° field of view each, sampling data every 30 seconds. It sends smartphone notifications when the bladder reaches a user-defined threshold. DFree (Triple W Japan, Japan) operates similarly, using A-mode ultrasound and proprietary algorithms to estimate bladder filling [12].

A recent development is EdgeFlow UW20 (Edgecare, Korea), which takes a different approach. Rather than A-mode ultrasound, UW20 utilizes a deep learning-based automatic bladder contour detection. This imaging-based approach enables quantitative bladder volume assessment comparable to that of conventional portable bladder scanners, while supporting continuous wearable monitoring. In this context, deep learning-based bladder detection and segmentation algorithms have demonstrated promising performance in bladder detection and segmentation [16-18]. Collectively, such developments may represent a potential bridge between the continuous monitoring capability of wearable devices and the quantitative accuracy required for clinical decision-making.

DIAGNOSTIC PERFORMANCE

Notification Accuracy for Bladder Fullness Detection

The key question for notification-based devices is straightforward: Do they reliably alert patients when their bladder is sufficiently full? Most of the available evidence for SENS-U comes from studies involving children with functional voiding dysfunction (FVD). At this time, we cannot assume these findings apply to other patient groups. Researchers evaluated SENS-U during urodynamic studies in 30 children with FVD. The device identified “full bladder” status with 90% sensitivity, and there was a strong Spearman correlation (rs=0.94) between device notifications and actual infused urine volume [14]. Another study tracked 68 children with FVD during daily activities, and SENS-U achieved 91% notification accuracy [15]. To date, this pediatric FVD group is the only context where moderatequality evidence supports these devices.

Historically, the URIKA monitor—the earlier version of SENS-U —showed average accuracy of 85% in detecting blad-der fullness. Notably, it performed perfectly in children under 10 (100% accuracy), but accuracy dropped to 71% among those older than 10 [10, 19]. This likely reflects changes in body composition with age.

Critical Concern: Poor Performance in Neurogenic Bladder

A recent study by Cai et al. [20] tested DFree in 40 children with neurogenic bladder (most with spina bifida). The results are striking: overall accuracy was only 46.2%, with sensitivity at 70%, but specificity falling to 13%. Such low specificity indicates a high false-positive rate, meaning the device often alarmed fullness when the bladder was not actually full.

For children who require lifelong catheterization, this is more than inconvenient. Excessive false positives can lead to unnecessary catheterizations, greater discomfort, increased caregiver burden, and a higher risk of UTIs. Based on these findings, current wearable devices are simply not suitable for patients with neurogenic bladder until their reliability in this population is established.

Comparison with Conventional Bladder Scanners

Conventional portable bladder scanners have demonstrated strong correlations with catheterized bladder volumes, with reported correlation coefficients ranging from r=0.79 to 0.95 across multiple validation studies. As a result, these devices are widely used in clinical practice to guide catheterization decisions and assess urinary retention.

However, direct head-to-head comparisons between wearable bladder monitoring devices and conventional scanners remain limited. As summarized in Table 1, most validation studies of wearable systems have employed urodynamic infusion or symptom-based endpoints as reference standards rather than catheterized volume measurements. In addition, wearable devices frequently report categorical bladder fullness or longitudinal trends rather than absolute volume estimates, which further complicates direct equivalence assessment.

Taken together, the currently available evidence does not support the interchangeability of wearable bladder monitoring devices and conventional bladder scanners when precise volume thresholds are required for clinical decision-making (Table 1).

CLINICAL APPLICATIONS BY POPULATION

Evidence Summary by Population

Table 2 summarizes the current evidence base for wearable bladder monitors for different clinical populations. Because available clinical evidence remains limited and heterogeneous, these classifications should be interpreted cautiously and primarily reflect expert interpretation rather than robust prospective trials.

Adult SCI and MS

Adult SCI and MS patients arguably have the greatest need for continuous bladder monitoring, but to date, there are no published prospective validation studies for these devices in these populations. Current European Association of Urology and International Continence Society guidelines do not mention wearable bladder monitors for neurogenic lower urinary tract dysfunction, reflecting the absence of supporting evidence [5, 6]. SCI patients with injuries above T6 are at real risk: AD can occur if the bladder becomes overdistended, resulting in sudden, severe hypertension, which can lead to stroke, seizures, or even death [4].

Pediatric studies have demonstrated that these devices can have extremely poor specificity (as low as 13%). Using unvalidated wearables in adults at risk of AD can result in missed true cases due to false negatives, or alarm fatigue from excessive false positives. Until prospective studies confirm that these devices are both highly sensitive and highly specific in adult SCI populations, they should not be used for patients at risk for AD.

It is not appropriate to extrapolate data from pediatric FVD to adult neurogenic patients. In Korea, while EdgeFlow UW20 recently received Ministry of Food and Drug Safety (MFDS) approval as a digital medical device, it has not yet been evaluated by Health Insurance Review & Assessment Service and no reimbursement pathway exists. The manufacturer has announced plans to expand clinical indications to neurogenic bladder populations including SCI and spina bifida, which would directly address this evidence gap.

Neobladder and Urinary Diversion

For patients with an orthotopic neobladder or continent urinary diversion, wearable A-mode ultrasound monitoring may be unreliable because of the altered reservoir anatomy. Neobladder reservoirs, constructed from bowel segments, often have irregular and variable shapes—unlike the smooth, round bladder geometry assumed by most device algorithms. Additionally, intestinal reservoirs may change position, and bowel gas frequently interferes with ultrasound signal transmission. Because these reservoirs lack a native detrusor muscle, acoustic reflection patterns may differ substantially from those expected by algorithms calibrated for the native bladder. However, direct clinical validation studies evaluating wearable bladder monitoring devices in neobladder populations are currently lacking. Therefore, these considerations should be interpreted as physiological rationale rather than evidence-based conclusions. For volume management in these patients, clinicians should rely on timed voiding strategies, conventional B-mode ultrasound imaging, and patient education.

REGULATORY STATUS AND COMMERCIAL AVAILABILITY

Wearable bladder devices exist in a regulatory gray area, and the rules vary depending on location and device (Table 3). SENS-U is CE-marked as a Class IIa medical device for pediatric use in Europe. DFree (Triple W Japan, Japan) is marketed as a consumer wellness product—the manufacturer explicitly states “DFree is not a medical device and is not intended to be used in the diagnosis, prevention, or treatment of diseases or medical conditions” [21]. As of November 2025, there is no 510(k) clearance for DFree in the U.S. Food and Drug Administration database. This raises significant questions about liability if clinicians use it for medical decision-making.

EdgeFlow UW20 received MFDS approval in April 2025 as a digital medical device—the first wearable bladder monitor with this classification in Korea [22]. Unlike DFree, it is regulated as a medical device with corresponding quality standards and manufacturer accountability. Unlike SENS-U, it employs Bmode imaging with deep learning rather than A-mode ultrasound. Clinical validation studies are currently underway at Severance Hospital, with plans to expand indications beyond postoperative retention to neurogenic bladder populations.

PROPOSED CLINICAL FRAMEWORK

Fig. 2 presents a framework—based on expert opinion—for when to consider wearable bladder monitors in practice. This is not an official guideline, as there is not enough evidence yet to support such a guideline. Instead, it combines current knowledge and expert recommendations to help clinicians make thoughtful, individualized decisions.

Fig. 2.

Conceptual framework for clinical use of wearable bladder monitoring devices in neurourology. The framework illustrates potential clinical decision pathways based on currently available evidence and expert interpretation. Wearable devices are proposed primarily as adjunct tools for monitoring bladder filling trends rather than replacements for conventional quantitative bladder scanners. Prospective studies are required to validate this framework by evaluating clinically relevant outcomes such as catheterization timing accuracy, urinary tract infection reduction, and prevention of bladder overdistension. BMI, body mass index; AD, autonomic dysreflexia; VD, voiding dysfunction.

The framework is built on several points: Wearables should supplement, not replace, standard quantitative assessments. Only use them in populations where validation data exist—at present, this is limited to pediatric FVD. Do not use wearables if the patient has altered bladder anatomy or position (such as a neobladder or major pelvic organ prolapse), or if their body habitus (body mass index over 35 kg/m²) makes ultrasound unreliable. For patients at risk of AD, stick to conventional bladder scanners—wearables are not validated for these cases. If care requires precise bladder volume thresholds, always confirm with a standard scanner.

DISCUSSION AND CONCLUSIONS

This review highlights a clear gap between the anticipated clinical potential of wearable bladder monitoring devices and the level of evidence currently available to support their routine use in neurourology. At present, published studies are limited in scale, often emphasize feasibility or usability, and frequently underrepresent the patient populations most likely to benefit from continuous bladder monitoring.

A critical distinction must be made between quantitative measurement of PVR volume and qualitative assessment of bladder filling trends. First-generation wearable devices based on A-mode ultrasound—such as SENS-U and DFree—are not designed to provide precise volumetric measurements [17, 18, 21-23]. Instead, they generate categorical estimates or longitudinal trends of bladder fullness. In this context, such devices may serve as adjunctive tools for monitoring voiding patterns or timing interventions, whereas quantitative clinical decisions continue to rely on conventional bladder scanners.

The reduced performance reported in neurogenic bladder populations (46% accuracy, 13% specificity) likely reflects inherent limitations of A-mode ultrasound technology [20]. Structural alterations commonly observed in neurogenic bladders— including trabeculation and heterogeneous wall thickness [1, 2]—can substantially affect ultrasound reflection patterns. In addition, functional changes such as altered bladder compliance further complicate signal interpretation by modifying pressure–volume relationships during filling. Together, these structural and functional abnormalities pose significant challenges for 1D signal-based algorithms that are not optimized to accommodate complex anatomical and biomechanical variability.

In parallel, regulatory status varies across currently available wearable bladder monitoring systems. Some devices have obtained medical device approval for specific indications, whereas others are marketed as wellness or symptom-awareness tools [21]. These distinctions should be considered when interpreting reported performance metrics and assessing clinical applicability, particularly in high-risk patient populations. Regulatory classification also has practical implications for clinical adoption: devices approved as medical devices must undergo formal clinical validation, quality control, and postmarket surveillance, whereas consumer wellness products may enter the market with fewer regulatory requirements. From an intellectual property perspective, medical device developers typically invest in proprietary clinical validation datasets and algorithm certification pipelines, whereas consumer wellness products may adopt more flexible, iterative development strategies with fewer documentation requirements. Regarding responsibilities, medical device manufacturers bear formal legal accountability for device-related adverse events and are obligated to conduct postmarket surveillance; conversely, when clinicians use consumer wellness products for medical decision-making, the responsibility for clinical outcomes shifts substantially toward the individual practitioner, creating potential liability concerns. As wearable monitoring technologies increasingly incorporate software and machine-learning-based algorithms, regulatory frameworks governing software updates and algorithm modifications may also influence their safe integration into clinical practice.

Advances in B-mode ultrasound combined with deep learning-based image analysis represent a promising technological evolution. Unlike first-generation wearable systems based on Amode ultrasound, B-mode imaging enables direct visualization of bladder contours and therefore has the potential to provide more accurate quantitative bladder volume estimation. Recent studies have demonstrated that convolutional neural networks can achieve accurate bladder segmentation and volume estimation across a wide range of bladder capacities [16-18]. A prospective study from Seoul National University Hospital reported that deep learning-based algorithms achieved more accurate bladder volume measurements compared with conventional methods across the entire volume range [18]. More recently, wearable ultrasound platforms incorporating flexible transducer arrays and machine learning-based bladder segmentation have demonstrated the feasibility of continuous quantitative bladder volume monitoring in experimental settings [16-18, 24-26]. These developments suggest that wearable bladder monitoring technology may be evolving from signal-based alert systems toward imaging-based wearable platforms capable of quantitative bladder volume assessment. In this context, the recent MFDS approval of the EdgeFlow UW20 device in April 2025 reflects an alternative approach that employs deep learning– based algorithms rather than A-mode signal analysis. Clinical validation studies for this device are ongoing.

Taken together, wearable ultrasound devices for bladder monitoring appear to be at a transitional stage. While first-generation A-mode devices have shown acceptable performance in selected pediatric FVD cohorts [14, 15, 19], their limitations in neurogenic bladder underscore the need for next-generation solutions. For patients with neurogenic bladder conditions— including SCI, MS, and spina bifida—the current evidence remains insufficient or concerning.

The conceptual framework presented in Fig. 2 is based on currently available evidence and expert interpretation rather than formal guideline recommendations. Future studies should prospectively evaluate the clinical utility of this framework by assessing whether wearable bladder monitoring can improve clinically meaningful outcomes. Potential study designs include prospective cohort studies or randomized trials comparing wearable monitoring–guided bladder management with conventional bladder scanning strategies. Relevant endpoints may include reduction in UTIs, improved adherence to CIC schedules, prevention of bladder overdistension or AD, and patientreported quality-of-life outcomes.

Future clinical trials in SCI and MS populations should employ clinically meaningful primary endpoints rather than surrogate measures alone. Recommended primary endpoints include: (1) reduction in symptomatic UTI rates, which directly reflects bladder management quality; (2) successful CIC adherence rate, defined as catheterization performed within the target bladder volume range (e.g., 300–400 mL); (3) incidence of AD episodes attributable to bladder overdistension; and (4) patient-reported quality of life using validated instruments such as the Qualiveen questionnaire. Secondary endpoints should include catheterization frequency optimization, reduction in emergency department visits for urological complications, and long-term preservation of upper urinary tract function. Trial designs should incorporate a minimum 6-month follow-up period to capture clinically relevant outcomes and should stratify participants by injury level (above vs. below T6) given the differential AD risk.

For patients at risk of AD, we propose minimum acceptable diagnostic thresholds based on the clinical consequences of device failure. Given that AD can be life-threatening, false negatives (missed overdistension) carry greater clinical risk than false positives (unnecessary catheterization). Therefore, we recommend a minimum sensitivity of ≥95% to ensure reliable detection of bladder overdistension before AD onset. While high specificity is desirable to minimize unnecessary catheterizations and alarm fatigue, a specificity threshold of ≥80% may be acceptable given the relative safety of precautionary catheterization compared to missed AD episodes. The positive predictive value should exceed 85% in the target population to maintain clinical utility. These thresholds are more stringent than those acceptable for FVD populations, where the consequences of false results are less severe. Until wearable devices demonstrate performance meeting these criteria in prospective validation studies involving adult SCI patients with injuries at T6 or above, conventional quantitative bladder scanning should remain the standard of care for this population.

At present, conventional quantitative bladder scanning remains the standard of care, particularly for patients at risk of AD. Future research priorities should include prospective validation of B-mode deep learning–based wearable systems in adult neurogenic populations, evaluation of clinically meaningful outcomes such as UTI rates, and head-to-head comparisons with conventional scanners, as highlighted by recent scoping and trial-design literature [13, 23]. As this field evolves rapidly, continued critical appraisal of emerging evidence will be essential before widespread clinical adoption can be recommended.