Machine learning applications to enhance patient specific care for urologic surgery.

PURPOSE: As computational power has improved over the past 20 years, the daily application of machine learning methods has become more prevalent in daily life. Additionally, there is increasing interest in the clinical application of machine learning techniques. We sought to review the current literature regarding machine learning applications for patient-specific urologic surgical care. METHODS: We performed a broad search of the current literature via the PubMed-Medline and Google Scholar databases up to Dec 2020. The search terms "urologic surgery" as well as "artificial intelligence", "machine learning", "neural network", and "automation" were used. RESULTS: The focus of machine learning applications for patient counseling is disease-specific. For stone disease, multiple studies focused on the prediction of stone-free rate based on preoperative characteristics of clinical and imaging data. For kidney cancer, many studies focused on advanced imaging analysis to predict renal mass pathology preoperatively. Machine learning applications in prostate cancer could provide for treatment counseling as well as prediction of disease-specific outcomes. Furthermore, for bladder cancer, the reviewed studies focus on staging via imaging, to better counsel patients towards neoadjuvant chemotherapy. Additionally, there have been many efforts on automatically segmenting and matching preoperative imaging with intraoperative anatomy. CONCLUSION: Machine learning techniques can be implemented to assist patient-centered surgical care and increase patient engagement within their decision-making processes. As data sets improve and expand, especially with the transition to large-scale EHR usage, these tools will improve in efficacy and be utilized more frequently.

Patient-specific, touch-based registration during robotic, image-guided partial nephrectomy.

Image-guidance during partial nephrectomy enables navigation within the operative field alongside a 3-dimensional roadmap of renal anatomy generated from patient-specific imaging. Once a process is performed by the human mind, the technology will allow standardization of the task for the benefit of all patients undergoing robot-assisted partial nephrectomy. Any surgeon will be able to visualize the kidney and key subsurface landmarks in real-time within a 3-dimensional simulation, with the goals of improving operative efficiency, decreasing surgical complications, and improving oncologic outcomes. For similar purposes, image-guidance has already been adopted as a standard of care in other surgical fields; we are now at the brink of this in urology. This review summarizes touch-based approaches to image-guidance during partial nephrectomy, as the technology begins to enter in vivo human evaluation. The processes of segmentation, localization, registration, and re-registration are all described with seamless integration into the da Vinci surgical system; this will facilitate clinical adoption sooner.

Racial Disparities in Diagnosis and Management of Benign Prostatic Hyperplasia: A Review.

PURPOSE OF REVIEW: We sought to perform a contemporary literature review highlighting the racial disparities which exists in the evaluation and management of benign prostatic hyperplasia (BPH). RECENT FINDINGS: Current literature suggests that racial disparities exist in the diagnosis of BPH and treatment lower urinary tract symptoms (LUTS). This is seen in the presentation and diagnosis of the disease as well as a difference in preventative care with discordant incidences of medical and surgical management among racial groups. The racial disparities that exist in the diagnosis and management of BPH and LUTS require further investigation to better identify the underlying causes. This will ultimately allow for continued improvement in care delivery and a more personalized approach in patient management.

MOSESTM Technology for Holmium Laser Enucleation of the Prostate: A Prospective Double-Blind Randomized Controlled Trial.

PURPOSE: Holmium laser enucleation of the prostate has proven to be efficacious and safe for the treatment of benign prostatic hyperplasia. New laser technologies, such as the MOSES™ pulse laser system, improve energy delivery and may improve operative times. We sought to prospectively evaluate holmium laser enucleation of the prostate using MOSES technology in a double-blind randomized controlled trial. MATERIALS AND METHODS: This is a single-center, prospective, double-blind, randomized controlled trial comparing holmium laser enucleation of the prostate using MOSES technology to holmium laser enucleation of the prostate. Patients were randomized in a 1:1 fashion. The study was powered to evaluate for a difference in operative time. Secondary end points included enucleation, morcellation, and hemostasis times, as well as blood loss, functional outcomes and complications 6 weeks postoperatively. RESULTS: A total of 60 patients were analyzed without difference in preoperative characteristics in either group (holmium laser enucleation of the prostate using MOSES technology: 30/60, 50%, holmium laser enucleation of the prostate: 30/60, 50%). Shorter total operative time was seen in the holmium laser enucleation of the prostate using MOSES technology group compared to the holmium laser enucleation of the prostate group (mean: 101 vs. 126 minutes, p <0.01). This difference remained significant on multiple linear regression. Additionally, the holmium laser enucleation of the prostate using MOSES technology group had shorter enucleation times (mean: 68 vs. 80 minutes, p=0.03), hemostasis time (mean: 18 vs. 29 minutes, p <0.01), and less blood loss (mean: -6.3 vs. -9.0%, p=0.03), measured by a smaller change in hematocrit postoperatively, compared to the traditional holmium laser enucleation of the prostate. There was no difference in functional or safety outcomes at followup. CONCLUSIONS: We report the results of a prospective, double-blind, randomized controlled trial comparing holmium laser enucleation of the prostate using MOSES technology to traditional holmium laser enucleation of the prostate. MOSES technology resulted in an improvement in operative time and a reduction in blood loss with comparable functional outcomes and complications compared to traditional holmium laser enucleation of the prostate.

Diagnostic Performance and Interreader Agreement of a Standardized MR Imaging Approach in the Prediction of Small Renal Mass Histology.

Purpose To assess the diagnostic performance and interreader agreement of a standardized diagnostic algorithm in determining the histologic type of small (≤4 cm) renal masses (SRMs) with multiparametric magnetic resonance (MR) imaging. Materials and Methods This single-center retrospective HIPAA-compliant institutional review board-approved study included 103 patients with 109 SRMs resected between December 2011 and July 2015. The requirement for informed consent was waived. Presurgical renal MR images were reviewed by seven radiologists with diverse experience. Eleven MR imaging features were assessed, and a standardized diagnostic algorithm was used to determine the most likely histologic diagnosis, which was compared with histopathology results after surgery. Interreader variability was tested with the Cohen κ statistic. Regression models using MR imaging features were used to predict the histopathologic diagnosis with 5% significance level. Results Clear cell renal cell carcinoma (RCC) and papillary RCC were diagnosed, with sensitivities of 85% (47 of 55) and 80% (20 of 25), respectively, and specificities of 76% (41 of 54) and 94% (79 of 84), respectively. Interreader agreement was moderate to substantial (clear cell RCC, κ = 0.58; papillary RCC, κ = 0.73). Signal intensity (SI) of the lesion on T2-weighted MR images and degree of contrast enhancement (CE) during the corticomedullary phase were independent predictors of clear cell RCC (SI odds ratio [OR]: 3.19; 95% confidence interval [CI]: 1.4, 7.1; P = .003; CE OR, 4.45; 95% CI: 1.8, 10.8; P < .001) and papillary RCC (CE OR, 0.053; 95% CI: 0.02, 0.2; P < .001), and both had substantial interreader agreement (SI, κ = 0.69; CE, κ = 0.71). Poorer performance was observed for chromophobe histology, oncocytomas, and minimal fat angiomyolipomas, (sensitivity range, 14%-67%; specificity range, 97%-99%), with fair to moderate interreader agreement (κ range = 0.23-0.43). Segmental enhancement inversion was an independent predictor of oncocytomas (OR, 16.21; 95% CI: 1.0, 275.4; P = .049), with moderate interreader agreement (κ = 0.49). Conclusion The proposed standardized MR imaging-based diagnostic algorithm had diagnostic accuracy of 81% (88 of 109) and 91% (99 of 109) in the diagnosis of clear cell RCC and papillary RCC, respectively, while achieving moderate to substantial interreader agreement among seven radiologists. © RSNA, 2018 Online supplemental material is available for this article.

Diagnostic Accuracy of Multiparametric Magnetic Resonance Imaging to Identify Clear Cell Renal Cell Carcinoma in cT1a Renal Masses.

PURPOSE: The detection of small renal masses is increasing with the use of cross-sectional imaging, although many incidental lesions have negligible metastatic potential. Among malignant masses clear cell renal cell carcinoma is the most prevalent and aggressive subtype. A method to identify such histology would aid in risk stratification. Our goal was to evaluate a likelihood scale for multiparametric magnetic resonance imaging in the diagnosis of clear cell histology. MATERIALS AND METHODS: We retrospectively reviewed the records of patients with cT1a masses who underwent magnetic resonance imaging and partial or radical nephrectomy from December 2011 to July 2015. Seven radiologists with different levels of experience who were blinded to final pathology findings independently reviewed studies based on a predefined algorithm. They applied a clear cell likelihood score, including 1-definitely not, 2-probably not, 3-equivocal, 4-probably and 5-definitely. Binary classification was used to determine the accuracy of clear cell vs all other histologies. Interobserver agreement was calculated with the weighted κ statistic. RESULTS: A total of 110 patients with 121 masses were identified. Mean tumor size was 2.4 cm and 50% of the lesions were clear cell. Defining clear cell as scores of 4 or greater demonstrated 78% sensitivity and 80% specificity while scores of 3 or greater showed 95% sensitivity and 58% specificity. Interobserver agreement was moderate to good with a mean κ of 0.53. CONCLUSIONS: A clear cell likelihood score used with magnetic resonance imaging can reasonably identify clear cell histology in small renal masses and may decrease the number of diagnostic renal mass biopsies. Standardization of imaging protocols and reporting criteria is needed to improve interobserver reliability.

"Stiction Syndrome": Non-Operative Management of Patients With Difficult AMS 700 Series Inflation.

INTRODUCTION: Static friction (stiction) is a mechanical phenomenon in which a state of increased resistance exists across a control valve mechanism. AIM: To present a strategy for non-operative management of inflatable penile prosthesis (IPP) cases with pump malfunction from pump valve stiction. METHODS: All patients had American Medical Systems (AMS; Minnetonka, MN, USA) 700 series Momentary Squeeze IPPs with transient pump malfunction owing to pump valve stiction after extended periods of device inactivity. MAIN OUTCOME MEASURES: Our evolving non-operative management experience with the "forced deflation" maneuver is described. This technique has successfully prevented the need for surgical pump replacement. Of patients with IPP who were instructed to inflate and deflate daily to prevent stiction recurrence, none have re-presented with difficult inflation. RESULTS: Of 306 patients receiving the AMS 700 series IPP at our institution from 2007 through 2015, 6 (1.9%) presented with difficulty activating the Momentary Squeeze pump (from 2011 through 2015). Four additional patients were referred from outside institutions with the same complaint. All patients (10 of 10, 100%) presented after a prolonged period of inactivity (minimum = 6 weeks) during which the IPP was not cycled and remained stagnant. Although the initial four patients (40%) underwent surgical exploration with pump mobilization and replacement, the six most recent patients (60%) were successfully instructed in the forced deflation maneuver in the office, which enabled the device to cycle normally thereafter. CONCLUSION: Device inactivity, for as little as 6 weeks, can predispose to Momentary Squeeze pump valve malfunction; and a regimen of daily IPP cycling could prevent stiction-related malfunction. Our findings should encourage practitioners to attempt conservative management of patients with "stiction syndrome" whenever possible, thereby avoiding unnecessary surgery. Kavoussi NL, Viers BR, VanDyke ME, et al. "Stiction Syndrome": Non-Operative Management of Patients With Difficult AMS 700 Series Inflation. J Sex Med 2017;14:1079-1083.

Synchronous Ipsilateral High Submuscular Placement of Prosthetic Balloons and Reservoirs.

INTRODUCTION: Synchronous ipsilateral high submuscular placement of artificial urinary sphincter (AUS) pressure-regulating balloons (PRBs) and inflatable penile prosthesis (IPP) reservoirs in a single submuscular tunnel is a novel strategy that could be advantageous for patients who have had major pelvic surgery. AIM: To report our initial experience with synchronous ipsilateral vs bilateral placement of AUS PRBs and IPP reservoirs in men undergoing implant surgery. METHODS: We retrospectively reviewed all patients undergoing synchronous AUS and IPP placement from 2007 through 2015 by a single surgeon at our tertiary center. Patients were stratified according to ipsilateral vs bilateral placement of the AUS PRB and IPP reservoir. MAIN OUTCOME MEASURES: Reoperation rates because of infectious or erosive complications and mechanical failure were assessed. RESULTS: Of the 968 implant surgeries during the study period, 47 men had synchronous device placement, of whom 17 (36%) underwent ipsilateral placement of the PRB and reservoir. During a median follow-up of 19 months (range = 1-84 months), reoperations were necessary in 12 of 47 (26%) and were similar between groups (ipsilateral, 5 of 17, 29%; bilateral, 7 of 30, 23%; P = .73). Most reoperations were due to AUS-related complications (10 of 12, 83%) and nearly all patients with reoperation (10 of 12, 83%) had compromised urethras (ie, prior urethral surgery, radiation, or prior AUS implantation). The most common indication for reintervention was cuff erosion (4 of 47, 9%), with no difference between groups (ipsilateral, 3 of 17, 18%; bilateral, 1 of 30, 3%; P = .13). CONCLUSION: Synchronous ipsilateral high submuscular placement of urologic prosthetic balloons could safely facilitate prosthetic surgery in patients with a history of major pelvic and inguinal surgery.

Preoperative Urine Culture Results Correlate Poorly With Bacteriology of Urologic Prosthetic Device Infections.

INTRODUCTION: Although preoperative negative urine culture results and treatment of urinary tract infections are generally advised before artificial urinary sphincter (AUS) and penile prosthesis (PP) surgery to prevent device infection, limited evidence exists to support this practice. AIM: To evaluate the relation between preoperative urine culture results and the bacteriology of prosthetic device infections. METHODS: Men undergoing AUS and/or PP placement at a tertiary referral center from 2007 through 2015 were analyzed. A total of 713 devices were implanted in 681 patients (337 AUSs in 314 patients and 376 PPs in 367 patients), of whom 259 (36%) did not have preoperative urine culture and were excluded. The remaining 454 patients received standard broad-spectrum perioperative antibiotics. Two patient groups were identified based on preoperative urine cultures: group 1 had negative urine culture results and group 2 had untreated asymptomatic positive urine culture results identified postoperatively. MAIN OUTCOME MEASURES: Device infection was diagnosed clinically and cultures obtained from the explanted device and tissue spaces were compared with preoperative urine culture results. RESULTS: Although multivariate analysis showed that patients undergoing AUS placement had a 4.5-fold greater risk of positive urine culture results (114 of 250, 45%) compared with those undergoing PP placement (36 of 204, 18%; P < .001), infection rates between device types were similar (8 of 250 for AUSs [3%] and 7 of 204 for PPs [3%]; P = .89). At a median follow-up of 15 months, device infection occurred in 15 of 454 devices (3%) implanted and no differences in infection rates were noted between urine culture groups (10 of 337 in group 1 [3.3%] and 5 of 117 in group 2 [4.3%]; P = .28). Remarkably, only 1 of 15 device infections (7%) had the same organism present at preoperative urine culture. CONCLUSIONS: Despite the finding that patients with AUS placement had a 4.5 times higher rate of positive urine culture results than patients with PP placement, preoperative urine culture results appeared to show little correlation with the bacteriology of prosthetic device infections.

Ablative Therapies for the Treatment of Small Renal Masses: a Review of Different Modalities and Outcomes.

The widespread utilization of abdominal imaging has led to an increase in incidentally detected small renal masses. Although partial nephrectomy is still considered the gold standard treatment for these masses, there are risks associated with surgical excision, potentially limiting treatment for older patients with multiple comorbidities. A variety of ablative techniques have developed over the past several decades, altering the management of small renal masses. It is likely that improvements in technology will only broaden the applications of ablative therapy. This article provides an update on the various ablative techniques and outcomes.

Automated Upper Tract Urothelial Carcinoma Tumor Segmentation During Ureteroscopy Using Computer Vision Techniques.

Introduction: Endoscopic tumor ablation of upper tract urothelial carcinoma (UTUC) allows for tumor control with the benefit of renal preservation but is impacted by intraoperative visibility. We sought to develop a computer vision model for real-time, automated segmentation of UTUC tumors to augment visualization during treatment. Materials and Methods: We collected 20 videos of endoscopic treatment of UTUC from two institutions. Frames from each video (N = 3387) were extracted and manually annotated to identify tumors and areas of ablated tumor. Three established computer vision models (U-Net, U-Net++, and UNext) were trained using these annotated frames and compared. Eighty percent of the data was used to train the models while 10% was used for both validation and testing. We evaluated the highest performing model for tumor and ablated tissue segmentation using a pixel-based analysis. The model and a video overlay depicting tumor segmentation were further evaluated intraoperatively. Results: All 20 videos (mean 36 ± 58 seconds) demonstrated tumor identification and 12 depicted areas of ablated tumor. The U-Net model demonstrated the best performance for segmentation of both tumors (area under the receiver operating curve [AUC-ROC] of 0.96) and areas of ablated tumor (AUC-ROC of 0.90). In addition, we implemented a working system to process real-time video feeds and overlay model predictions intraoperatively. The model was able to annotate new videos at 15 frames per second. Conclusions: Computer vision models demonstrate excellent real-time performance for automated upper tract urothelial tumor segmentation during ureteroscopy.

Towards navigation in endoscopic kidney surgery based on preoperative imaging.

Endoscopic renal surgeries have high re-operation rates, particularly for lower volume surgeons. Due to the limited field and depth of view of current endoscopes, mentally mapping preoperative computed tomography (CT) images of patient anatomy to the surgical field is challenging. The inability to completely navigate the intrarenal collecting system leads to missed kidney stones and tumors, subsequently raising recurrence rates. A guidance system is proposed to estimate the endoscope positions within the CT to reduce re-operation rates. A Structure from Motion algorithm is used to reconstruct the kidney collecting system from the endoscope videos. In addition, the kidney collecting system is segmented from CT scans using 3D U-Net to create a 3D model. The two collecting system representations can then be registered to provide information on the relative endoscope position. Correct reconstruction and localization of intrarenal anatomy and endoscope position is demonstrated. Furthermore, a 3D map is created supported by the RGB endoscope images to reduce the burden of mental mapping during surgery. The proposed reconstruction pipeline has been validated for guidance. It can reduce the mental burden for surgeons and is a step towards the long-term goal of reducing re-operation rates in kidney stone surgery.

ASSIST-U: A system for segmentation and image style transfer for ureteroscopy.

Kidney stones require surgical removal when they grow too large to be broken up externally or to pass on their own. Upper tract urothelial carcinoma is also sometimes treated endoscopically in a similar procedure. These surgeries are difficult, particularly for trainees who often miss tumours, stones or stone fragments, requiring re-operation. Furthermore, there are no patient-specific simulators to facilitate training or standardized visualization tools for ureteroscopy despite its high prevalence. Here a system ASSIST-U is proposed to create realistic ureteroscopy images and videos solely using preoperative computerized tomography (CT) images to address these unmet needs. A 3D UNet model is trained to automatically segment CT images and construct 3D surfaces. These surfaces are then skeletonized for rendering. Finally, a style transfer model is trained using contrastive unpaired translation (CUT) to synthesize realistic ureteroscopy images. Cross validation on the CT segmentation model achieved a Dice score of 0.853 ± 0.084. CUT style transfer produced visually plausible images; the kernel inception distance to real ureteroscopy images was reduced from 0.198 (rendered) to 0.089 (synthesized). The entire pipeline from CT to synthesized ureteroscopy is also qualitatively demonstrated. The proposed ASSIST-U system shows promise for aiding surgeons in the visualization of kidney ureteroscopy.

Intraoperative gaze guidance with mixed reality.

Efficient communication and collaboration are essential in the operating room for successful and safe surgery. While many technologies are improving various aspects of surgery, communication between attending surgeons, residents, and surgical teams is still limited to verbal interactions that are prone to misunderstandings. Novel modes of communication can increase speed and accuracy, and transform operating rooms. A mixed reality (MR) based gaze sharing application on Microsoft HoloLens 2 headset that can help expert surgeons indicate specific regions, communicate with decreased verbal effort, and guide novices throughout an operation is presented. The utility of the application is tested with a user study of endoscopic kidney stone localization completed by urology experts and novice surgeons. Improvement is observed in the NASA task load index surveys (up to 25.23%), in the success rate of the task (6.98% increase in localized stone percentage), and in gaze analyses (up to 31.99%). The proposed application shows promise in both operating room applications and surgical training tasks.

Are 3D Image Guidance Systems Ready for Use? A Comparative Analysis of 3D Image Guidance Implementations in Minimally Invasive Partial Nephrectomy.

Introduction: Three-dimensional image-guided surgical (3D-IGS) systems for minimally invasive partial nephrectomy (MIPN) can potentially improve the efficiency and accuracy of intraoperative anatomical localization and tumor resection. This review seeks to analyze the current state of research regarding 3D-IGS, including the evaluation of clinical outcomes, system functionality, and qualitative insights regarding 3D-IGS's impact on surgical procedures. Methods: We have systematically reviewed the clinical literature pertaining to 3D-IGS deployed for MIPN. For inclusion, studies must produce a patient-specific 3D anatomical model from two-dimensional imaging. Data extracted from the studies include clinical results, registration (alignment of the 3D model to the surgical scene) method used, limitations, and data types reported. A subset of studies was qualitatively analyzed through an inductive coding approach to identify major themes and subthemes across the studies. Results: Twenty-five studies were included in the review. Eight (32%) studies reported clinical results that point to 3D-IGS improving multiple surgical outcomes. Manual registration was the most utilized (48%). Soft tissue deformation was the most cited limitation among the included studies. Many studies reported qualitative statements regarding surgeon accuracy improvement, but quantitative surgeon accuracy data were not reported. During the qualitative analysis, six major themes emerged across the nine applicable studies. They are as follows: 3D-IGS is necessary, 3D-IGS improved surgical outcomes, researcher/surgeon confidence in 3D-IGS system, enhanced surgeon ability/accuracy, anatomical explanation for qualitative assessment, and claims without data or reference to support. Conclusions: Currently, clinical outcomes are the main source of quantitative data available to point to 3D-IGS's efficacy. However, the literature qualitatively suggests the benefit of accurate 3D-IGS for robotic partial nephrectomy.

A Safe Framework for Quantitative In Vivo Human Evaluation of Image Guidance.

Goal: We present a new framework for in vivo image guidance evaluation and provide a case study on robotic partial nephrectomy. Methods: This framework (called the "bystander protocol") involves two surgeons, one who solely performs the therapeutic process without image guidance, and another who solely periodically collects data to evaluate image guidance. This isolates the evaluation from the therapy, so that in-development image guidance systems can be tested without risk of negatively impacting the standard of care. We provide a case study applying this protocol in clinical cases during robotic partial nephrectomy surgery. Results: The bystander protocol was performed successfully in 6 patient cases. We find average lesion centroid localization error with our IGS system to be 6.5 mm in vivo compared to our prior result of 3.0 mm in phantoms. Conclusions: The bystander protocol is a safe, effective method for testing in-development image guidance systems in human subjects.

Artificial intelligence model for automated surgical instrument detection and counting: an experimental proof-of-concept study.

BACKGROUND: Retained surgical items (RSI) are preventable events that pose a significant risk to patient safety. Current strategies for preventing RSIs rely heavily on manual instrument counting methods, which are prone to human error. This study evaluates the feasibility and performance of a deep learning-based computer vision model for automated surgical tool detection and counting. METHODS: A novel dataset of 1,004 images containing 13,213 surgical tools across 11 categories was developed. The dataset was split into training, validation, and test sets at a 60:20:20 ratio. An artificial intelligence (AI) model was trained on the dataset, and the model's performance was evaluated using standard object detection metrics, including precision and recall. To simulate a real-world surgical setting, model performance was also evaluated in a dynamic surgical video of instruments being moved in real-time. RESULTS: The model demonstrated high precision (98.5%) and recall (99.9%) in distinguishing surgical tools from the background. It also exhibited excellent performance in differentiating between various surgical tools, with precision ranging from 94.0 to 100% and recall ranging from 97.1 to 100% across 11 tool categories. The model maintained strong performance on a subset of test images containing overlapping tools (precision range: 89.6-100%, and recall range 97.2-98.2%). In a real-time surgical video analysis, the model maintained a correct surgical tool count in all non-transition frames, with a median inference speed of 40.4 frames per second (interquartile range: 4.9). CONCLUSION: This study demonstrates that using a deep learning-based computer vision model for automated surgical tool detection and counting is feasible. The model's high precision and real-time inference capabilities highlight its potential to serve as an AI safeguard to potentially improve patient safety and reduce manual burden on surgical staff. Further validation in clinical settings is warranted.

Machine Learning Models to Predict Kidney Stone Recurrence Using 24 Hour Urine Testing and Electronic Health Record-Derived Features.

Objective: To assess the accuracy of machine learning models in predicting kidney stone recurrence using variables extracted from the electronic health record (EHR). Methods: We trained three separate machine learning (ML) models (least absolute shrinkage and selection operator regression [LASSO], random forest [RF], and gradient boosted decision tree [XGBoost] to predict 2-year and 5-year symptomatic kidney stone recurrence from electronic health-record (EHR) derived features and 24H urine data (n = 1231). ML models were compared to logistic regression [LR]. A manual, retrospective review was performed to evaluate for a symptomatic stone event, defined as pain, acute kidney injury or recurrent infections attributed to a kidney stone identified in the clinic or the emergency department, or for any stone requiring surgical treatment. We evaluated performance using area under the receiver operating curve (AUC-ROC) and identified important features for each model. Results: The 2- and 5- year symptomatic stone recurrence rates were 25% and 31%, respectively. The LASSO model performed best for symptomatic stone recurrence prediction (2-yr AUC: 0.62, 5-yr AUC: 0.63). Other models demonstrated modest overall performance at 2- and 5-years: LR (0.585, 0.618), RF (0.570, 0.608), and XGBoost (0.580, 0.621). Patient age was the only feature in the top 5 features of every model. Additionally, the LASSO model prioritized BMI and history of gout for prediction. Conclusions: Throughout our cohorts, ML models demonstrated comparable results to that of LR, with the LASSO model outperforming all other models. Further model testing should evaluate the utility of 24H urine features in model structure.