Fatemi, Mostafa

ABSTRACT Background Interpretation of urodynamic studies (UDS) is central to diagnosing lower urinary tract dysfunction, but it is often complex, time‐consuming, and subject to interobserver variability. Recent advances in artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), have enabled automated analysis of multichannel pressure–flow data and may improve objectivity and efficiency in UDS interpretation. Objective To critically review current applications of ML and DL for interpretation of multichannel UDS data and to evaluate their performance, clinical targets, and limitations. Methods A literature search of PubMed, Scopus, and Web of Science was performed through September 2, 2025, without language or date restrictions. A narrative review was conducted of peer‐reviewed studies applying ML or DL techniques to invasive UDS signals, including vesical, abdominal, and detrusor pressures as well as urinary flow. Studies limited to uroflowmetry or non‐UDS inputs were excluded. Model design, input features, validation strategies, and clinical outcomes were systematically assessed. Results A total of 12 studies met inclusion criteria. AI models were applied to a range of UDS tasks, including detection of detrusor overactivity, classification of bladder outlet obstruction and detrusor underactivity, multi‐feature UDS pattern recognition, severity grading, and real‐time event detection. Reported diagnostic performance was generally high, with many models achieving accuracies or AUCs between 80% and 95% for primary outcomes. However, study designs were heterogeneous, most datasets were retrospective and single‐center, and external validation was limited. Conclusions ML and DL approaches can extract clinically meaningful information from multichannel UDS recordings and demonstrate strong technical performance across several diagnostic tasks. However, variability in methodology, limited generalizability, and lack of prospective validation currently limit clinical adoption. AI‐based UDS interpretation remains an emerging research area requiring standardized data, robust validation, and integration with clinical workflows.

Introduction/Background: Peripheral arterial disease (PAD) exhibits reduced hyperemic response in the affected legs following stimuli such as pressure-cuff-induced occlusion or exercise. Therefore, an analysis of the perfusion changes in the calf muscle using a widely available imaging modality such as ultrasound could serve as a complementary tool to existing methods such as ankle-brachial index (ABI) for PAD detection. Research Questions/Hypothesis: This study evaluates the utility of ultrasound perfusion imaging and estimation technique in the calf muscle to differentiate PAD affected legs based on their flow response to external pressure and exercise. Methods/Approach: Data acquisition was performed using a research ultrasound system (Vantage NXT, Verasonics) with an ultrasound probe secured inside a probe fixation device and positioned to the calf muscle of the subject. Two separate studies were performed on each subject: 1) A flow variations response to external pressure study, 2) a flow variation response to exercise. For the pressure study, a pressure cuff was wrapped around the subject’s thigh and inflated during the study using an automatic rapid cuff inflation device. Data was continuously acquired for 6 minutes including 1 minute of baseline, 3 minutes of pressure cuff occlusion, 2 minutes of post-occlusion acquisition. The exercise study consisted of data acquisition in the seated position with 1 minute of baseline, 30 repetitions of plantar flexion using a calf training device, and 2 minutes of post-exercise acquisition. The acquired data were then clutter-filtered and denoised for perfusion visualization. Four metrics including post-occlusion/post-exercise to baseline flow intensity/density variation (POBFIV, PEBFIV, POBFDV, PEBFDV) were obtained for analysis. Results/Data: Figure 1 shows perfusion imaging results during the pressure and the exercise studies for two example legs, one healthy and one of a clinically diagnosed PAD patient (mild disease), respectively. Percentage-wise POBFIV, PEBFIV, POBFDV, and PEBFDV values were 44%, 57%, 197%, and 2231% for the healthy leg and 12%, 27%, 63%, and 181% for the PAD leg, respectively, shown in Table 1. Conclusion: The proposed method shows promise in differentiating PAD-affected legs from healthy individual’s through evaluating the hyperemic response to external pressure and exercise. Statistical analysis of the results for a cohort of healthy and PAD-affected legs will constitute future work.

Introduction: The ankle-brachial index (ABI) is the most common method for detecting peripheral arterial disease (PAD), but its sensitivity is limited, especially in patients with poorly compressible vessels. This study proposes a novel ultrasound-based approach for PAD diagnosis by analyzing calf muscle perfusion responses to pressure-cuff occlusion and exercise, interpreted using machine learning (ML) techniques. Methods: Under an IRB approved protocol and signed IRB-approved written consent, 48 participants, 43 symptomatic with abnormal ABI and 5 healthy volunteers were recruited for this study. High-frame-rate ultrasound B-mode image data (500 frames/s) were collected from 64 legs and categorized into two classes based on clinical diagnosis: Abnormal ABI (54 legs) and Healthy (10 legs). Legs with normal ABI from symptomatic patients were excluded to avoid diagnostic ambiguity, as these cases may lie within the diagnostic gray zone for PAD. Each leg underwent a 9-minute scanning protocol comprising 1 minute of baseline, 3 minutes of pressure cuff occlusion (using an automatic rapid inflation/deflation device), 2 minutes post-occlusion, 1 minute of plantar flexion exercise, and 2 minutes post-exercise. The ultrasound data were processed using singular-value decomposition (SVD) clutter filtering and noise equalization to generate Power-Doppler perfusion images at 2 frames per second. The PAD prediction pipeline involved data augmentation to address class imbalance and limited sample size, followed by histogram-based thresholding to extract blood pixels—intensity thresholding was avoided due to variability in pixel intensities. Blood pixel counts were calculated over time and used to train nine machine learning classification models (Figure 1). Results: The best-performing ML classification model achieved an accuracy of 77%, with the median accuracy across all models ranging between 70% and 72%. Notably, an ensemble model that combined Logistic Regression, Gradient Boosting, and Support Vector Machines using a majority voting strategy yielded the highest accuracy of 81% (Table 1). Conclusion: We proposed a novel ultrasound-based perfusion estimation method for diagnosing peripheral arterial disease (PAD), with data analyzed using ML approach. Although this is a feasibility study, the observed diagnostic accuracy exceeding 80% is promising. With further research and larger datasets, this approach has the potential to become a valuable tool for PAD.

ABSTRACT Background Urodynamic studies (UDS) are integral to the evaluation and management of lower urinary tract dysfunction. However, artifacts are frequently encountered, affecting both clinical interpretation and research reliability. Objective This expert narrative review examines the prevalence, classification, and clinical relevance of artifacts encountered during UDS, based on findings from the existing literature. Methods A focused literature review was conducted, analyzing studies that reported data on the sources and impact of UDS artifacts, particularly from the patient perspective. Emphasis was placed on identifying both technical and patient‐related contributors to artifact generation. Results Artifacts are common in UDS despite adherence to Good Urodynamic Practices recommended by the International Continence Society. Technical artifacts may stem from equipment‐related issues such as calibration errors, catheter displacement, and device variability. Patient‐related artifacts can arise from physical movement or test‐induced detrusor overactivity. These artifacts may lead to misinterpretation, misdiagnosis, and inappropriate interventions in clinical practice. Furthermore, they may affect the validity of research findings related to diagnostic tools for lower urinary tract symptoms. Conclusion Identifying and addressing UDS artifacts during the procedure or in post‐test interpretation is critical for accurate clinical decision‐making and reliable research outcomes. Efforts to minimize artifacts should be prioritized to improve the quality and utility of UDS.

A variety of advanced shear wave elastography techniques require the measurement of frequency-resolved phase velocity to properly characterize wave signals where factors such as tissue viscosity or guided wave effects are present. However, when applying these techniques in a dynamic environment where extrinsic motion cannot be limited, nonlinear and nonstationary trends can be expected in the motion signal used for phase velocity estimation. Because such extrinsic motion signals can contaminate a broad range of frequencies, they can produce significant errors in the phase velocity estimates. In this study, we propose a spatiotemporal detrending approach for elastography wave motion signals built from a 2D Whittaker smoother. This smoother has its origins in nonparametric regression and is well suited for fitting nonlinear curves of arbitrary shape. The resulting detrending filter was applied to Lamb wave signals from clinical ultrasound bladder vibrometry measurements. We found that the proposed detrending filter allowed for phase velocity estimates to be made in acquisitions that are otherwise be too corrupted by large motion. This works was supported by a grant from the NIH (DK131685)