Machine learning provides an accurate prognostication model for refractory overactive bladder treatment response and is noninferior to human experts

How Can We Show That Artificial Intelligence May Improve Our Assessment and Management of Lower Urinary Tract Dysfunctions?-ICI-RS 2024

AIMS

The integration of artificial intelligence (AI) into functional urology management must be assessed for its clinical utility, but hopefully will change, perhaps to revolutionize the way LUTD and other conditions are assessed, the aim being to offer patients more rapid and effective management which enhances patient outcomes. The aim of this proposal, discussed at the ICI-RS annual meeting, is to evaluate the available evidence on AI and the way it might change the approach to urodynamic (UDS) diagnoses, including overactive bladder syndrome (OAB), and perhaps other LUTDs such as bladder outflow obstruction.

METHODS

A compendium of discussion based on the current evidence related to AI and its potential applications in UDS and OAB.

RESULTS

AI-powered diagnostic tools are being developed to analyze complex datasets from urodynamic studies, imaging, and other diagnostic tests. AI systems can leverage large volumes of clinical data to recommend personalized treatment plans based on individual patient profiles to optimize surgical procedures, enhance diagnostic precision, tailor the therapy, reduce the risk of complications, and improve outcomes. In the future, AI will be able to provide tailored counseling regarding the outcomes and potential side effects of drugs and procedures to a given patient.

CONCLUSION

AI's role in functional urology has been poorly investigated, and its implementation across several areas may improve clinical care and the pathophysiological understanding of functional urologic conditions.

Current applications and future perspectives of artificial intelligence in functional urology and neurourology: how far can we get?

In the last few years, the scientific community has seen an increasing interest towards the potential applications of artificial intelligence in medicine and healthcare. In this context, urology represents an area of rapid development, particularly in uro-oncology, where a wide range of applications has focused on prostate cancer diagnosis. Other urological branches are also starting to explore the potential advantages of AI in the diagnostic and therapeutic process, and functional urology and neurourology are among them. Although the experiences in this sense have been quite limited so far, some AI applications have already started to show potential benefits, especially for urodynamic and imaging interpretation, as well as for the development of AI-based predictive models for treatment response. A few experiences on the use of ChatGPT to answer questions on functional urology and neurourology topics have also been reported. Conversely, AI applications in functional urology surgery remain largely unexplored. This paper provides a critical overview of the current evidence on this topic, highlighting the potential benefits for the diagnostic workflow, therapeutic evaluation and surgical training, as well as the current limitations that need to be addressed to enable the integration of this tools in the clinical practice in the future.

External Validation Demonstrates Machine Learning Models Outperform Human Experts in Prediction of Objective and Patient-reported Overactive Bladder Treatment Outcomes

OBJECTIVE

To predict treatment response for overactive bladder (OAB) for a specific patient remains elusive. We sought to develop accurate models using machine learning for prediction of objective and patient-reported treatment response to intravesical botulinum toxin (OBTX-A) injection. We sought to validate the models in a challenging setting using an external dataset of a markedly different patient cohort and dosing regimen. We hypothesized the model would outperform human experts and top available algorithms.

METHODS

Algorithms using "operator splitting" designed for accuracy and efficiency even in small training datasets with variable completeness, were trained to predict objective response and patient-reported symptomatic improvement using the ROSETTA trial cohort and validated using the ABC trial cohort of patients who underwent OBTX-A. Areas under the curve (AUC) of algorithms were compared to the top publicly-available machine-learning classifier XGBoost, logistic regression with cross validation, and human expert predictions in the external validation set.

RESULTS

In the validation set, the operator splitting neural network had AUC of 0.66 and outperformed XGBoost with DART (top available machine-learning classifier, AUC: 0.58), logistic regression (AUC 0.55), and human experts (AUC 0.47-0.53) for prediction of clinical responder status. It was similarly accurate in prediction of patient subjective improvement in symptoms following OBTX-A (AUC: 0.64), again outperforming other algorithms and human experts (AUC 0.41-0.62).

CONCLUSION

The neural network outperformed human experts and other machine-learning approaches in prediction of objective and patient-reported OBTX-A outcomes for OAB in a challenging independent validation cohort. Clinical implementation could improve counseling and treatment selection.

Applications of machine learning in urodynamics: A narrative review

BACKGROUND

Machine learning algorithms as a research tool, including traditional machine learning and deep learning, are increasingly applied to the field of urodynamics. However, no studies have evaluated how to select appropriate algorithm models for different urodynamic research tasks.

METHODS

We undertook a narrative review evaluating how the published literature reports the applications of machine learning in urodynamics. We searched PubMed up to December 2023, limited to the English language. We selected the following search terms: artificial intelligence, machine learning, deep learning, urodynamics, and lower urinary tract symptoms. We identified three domains for assessment in advance of commencing the review. These were the applications of urodynamic studies examination, applications of diagnoses of dysfunction related to urodynamics, and applications of prognosis prediction.

RESULTS

The machine learning algorithm applied in the field of urodynamics can be mainly divided into three aspects, which are urodynamic examination, diagnosis of urinary tract dysfunction and prediction of the efficacy of various treatment methods. Most of these studies were single-center retrospective studies, lacking external validation, requiring further validation of model generalization ability, and insufficient sample size. The relevant research in this field is still in the preliminary exploration stage; there are few high-quality multi-center clinical studies, and the performance of various models still needs to be further optimized, and there is still a distance from clinical application.

CONCLUSIONS

At present, there is no research to summarize and analyze the machine learning algorithms applied in the field of urodynamics. The purpose of this review is to summarize and classify the machine learning algorithms applied in this field and to guide researchers to select the appropriate algorithm model for different task requirements to achieve the best results.

Can we use machine learning to improve the interpretation and application of urodynamic data?: ICI-RS 2023

INTRODUCTION

A "Think Tank" at the International Consultation on Incontinence-Research Society meeting held in Bristol, United Kingdom in June 2023 considered the progress and promise of machine learning (ML) applied to urodynamic data.

METHODS

Examples of the use of ML applied to data from uroflowmetry, pressure flow studies and imaging were presented. The advantages and limitations of ML were considered. Recommendations made during the subsequent debate for research studies were recorded.

RESULTS

ML analysis holds great promise for the kind of data generated in urodynamic studies. To date, ML techniques have not yet achieved sufficient accuracy for routine diagnostic application. Potential approaches that can improve the use of ML were agreed and research questions were proposed.

CONCLUSIONS

ML is well suited to the analysis of urodynamic data, but results to date have not achieved clinical utility. It is considered likely that further research can improve the analysis of the large, multifactorial data sets generated by urodynamic clinics, and improve to some extent data pattern recognition that is currently subject to observer error and artefactual noise.

Utility of Machine Learning in the Management of Normal Pressure Hydrocephalus: A Systematic Review

BACKGROUND

In the past decade, many machine learning (ML) models have been used in the management of normal pressure hydrocephalus (NPH). This study aims at systematically reviewing those ML models.

METHODS

The PubMed, Embase, and Web of Science databases were searched for studies reporting applications of ML in NPH. Quality assessment was performed using Prediction model Risk Of Bias ASsessment Tool (PROBAST) and Transparent Reporting of a multivariable predication model for Individual Prognosis Or Diagnosis (TRIPOD) adherence reporting guidelines, and statistical analysis was performed with the level of significance of <0.05.

RESULTS

A total of 22 studies with 53 models were included in the review, of which the convolutional neural network was the most used model. Inputs used to train various models included clinical features, computed tomography scan, magnetic resonance imaging, intracranial pulse waveform characteristics, and perfusion infusion. The overall mean accuracy of the models was 77% (highest for the convolutional neural network, 98%, while lowest for decision tree, 55%; P = 0.176). There was a statistically significant difference in the accuracy and area under the curve of diagnostic and interventional models (accuracy: 83.4% vs. 69.4%, area under the curve: 0.882 vs. 0.729; P < 0.001). Overall, 59.09% (n = 13) and 81.82% (n = 18) of the studies had high-risk bias and high-applicability, respectively, on PROBAST assessment; however, only 55.15% of the studies adhered to the TRIPOD statement.

CONCLUSIONS

Though highly accurate, there are many challenges to current ML models necessitating the need to standardize the ML models to enable comparison across the studies and enhance the NPH decision-making and care.

A treatment prediction strategy for overactive bladder using a machine learning algorithm that utilized data from the FAITH study

AIMS

To use machine learning algorithms to develop a model to accurately predict treatment responses to mirabegron or antimuscarinic agents in patients with overactive bladder (OAB), using real-world data from the FAITH registry (NCT03572231).

METHODS

The FAITH registry data included patients who had been diagnosed with OAB symptoms for at least 3 months and were due to initiate monotherapy with mirabegron or any antimuscarinic. For the development of the machine learning model, data from patients were included if they had completed the 183-day study period, had data for all timepoints and had completed the overactive bladder symptom scores (OABSS) at baseline and end of study. The primary outcome of the study was a composite outcome combining efficacy, persistence, and safety outcomes. Treatment was deemed "more effective" if the composite outcome criteria for "successful," "no treatment change," and "safe" were met, otherwise treatment was deemed "less effective." To explore the composite algorithm, a total of 14 clinical risk factors were included in the initial data set and a 10-fold cross-validation procedure was performed. A range of machine learning models were evaluated to determine the most effective algorithm.

RESULTS

In total, data from 396 patients were included (266 [67.2%] treated with mirabegron and 130 [32.8%] treated with an antimuscarinic). Of these, 138 (34.8%) were in the "more effective" group and 258 (65.2%) were in the "less effective" group. The groups were comparable in terms of their characteristic distributions across patient age, sex, body mass index, and Charlson Comorbidity Index. Of the six models initially selected and tested, the decision tree (C5.0) model was chosen for further optimization, and the receiver operating characteristic of the final optimized model had an area under the curve result of 0.70 (95% confidence interval: 0.54-0.85) when 15 was used for the min n parameter.

CONCLUSIONS

This study successfully created a simple, rapid, and easy-to-use interface that could be further refined to produce a valuable educational or clinical decision-making aid.

Neural networks outperform expert humans in predicting patient impressions of symptomatic improvement following overactive bladder treatment

INTRODUCTION AND HYPOTHESIS

The objective was to accurately predict patient-centered subjective outcomes following the overactive bladder (OAB) treatments OnabotulinumtoxinA (OBTX-A) injection and sacral neuromodulation (SNM) using a neural network-based machine-learning approach. In the context of treatments designed to improve quality of life, a patient's perception of improvement should be the gold standard outcome measure.

METHODS

Cutting-edge neural network-based algorithms using reproducing kernel techniques were trained to predict patient-reported improvements in urinary leakage and bladder function as assessed by Patient Global Impression of Improvement score using the ROSETTA trial datasets. Blinded expert urologists provided with the same variables also predicted outcomes. Receiver operating characteristic curves and areas under the curve were generated for algorithm and human expert predictions in an out-of-sample holdout dataset.

RESULTS

Algorithms demonstrated excellent accuracy in predicting patient subjective improvement in urinary leakage (OBTX-A: AUC 0.75; SNM: 0.80). Similarly, algorithms accurately predicted patient subjective improvement in bladder function (OBTX-A: AUC 0.86; SNM: 0.96). The top-performing algorithms outcompeted human experts across outcome measures.

CONCLUSIONS

Novel neural network-based machine-learning algorithms accurately predicted OBTX-A and SNM patient subjective outcomes, and generally outcompeted expert humans. Subtle aspects of the physician-patient interaction remain uncomputable, and thus the machine-learning approach may serve as an aid, rather than as an alternative, to human interaction and clinical judgment.