Development of a Deep Learning-Based Predictive Model for Improvement after Holmium Laser Enucleation of the Prostate According to Detrusor Contractility
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
Purpose: Predicting improvements in voiding symptoms following deobstructive surgery for male lower urinary tract symptoms/benign prostatic hyperplasia (LUTS/BPH) is challenging when detrusor contractility is impaired. This study aimed to develop an artificial intelligence model that predicts symptom improvement after holmium laser enucleation of the prostate (HoLEP), focusing on changes in maximum flow rate (MFR) and voiding efficiency (VE) 1-month postsurgery. Methods: We reviewed 1,933 patients who underwent HoLEP at Samsung Medical Center from July 2008 to January 2024. The study employed a deep neural network (DNN) for multiclass classification to predict changes in MFR and VE, each divided into 3 categories. For comparison, additional machine learning (ML) models such as extreme gradient boosting, random forest classification, and support vector machine were utilized. To address class imbalance, we applied the least squares method and multitask learning. Results: A total of 1,142 patients with complete data were included in the study, with 992 allocated for model training and 150 for external validation. In predicting MFR, the DNN achieved a microaverage area under the receiver operating characteristic curve (AUC) of 0.884±0.006, sensitivity of 0.783±0.020, and specificity of 0.891±0.010. For VE prediction, the microaverage AUC was 0.817±0.007, with sensitivity and specificity values of 0.660±0.014 and 0.830±0.007, respectively. These results indicate that the DNN’s predictive performance was superior to that of other ML models. Conclusions: The DNN model provides detailed and accurate predictions for recovery after HoLEP, providing valuable insights for clinicians managing patients with LUTS/BPH.
Abstract
Purpose: Predicting improvements in voiding symptoms following deobstructive surgery for male lower urinary tract symptoms/benign prostatic hyperplasia (LUTS/BPH) is challenging when detrusor contractility is impaired. This study aimed to develop an artificial intelligence model that predicts symptom improvement after holmium laser enucleation of the prostate (HoLEP), focusing on changes in maximum flow rate (MFR) and voiding efficiency (VE) 1-month postsurgery. Methods: We reviewed 1,933 patients who underwent HoLEP at Samsung Medical Center from July 2008 to January 2024. The study employed a deep neural network (DNN) for multiclass classification to predict changes in MFR and VE, each divided into 3 categories. For comparison, additional machine learning (ML) models such as extreme gradient boosting, random forest classification, and support vector machine were utilized. To address class imbalance, we applied the least squares method and multitask learning. Results: A total of 1,142 patients with complete data were included in the study, with 992 allocated for model training and 150 for external validation. In predicting MFR, the DNN achieved a microaverage area under the receiver operating characteristic curve (AUC) of 0.884±0.006, sensitivity of 0.783±0.020, and specificity of 0.891±0.010. For VE prediction, the microaverage AUC was 0.817±0.007, with sensitivity and specificity values of 0.660±0.014 and 0.830±0.007, respectively. These results indicate that the DNN’s predictive performance was superior to that of other ML models. Conclusions: The DNN model provides detailed and accurate predictions for recovery after HoLEP, providing valuable insights for clinicians managing patients with LUTS/BPH.
참고문헌 (25)
- Yoo TK, "The prevalence of lower urinary tract symptoms in population aged 40years or over, in South Korea", Investig Clin Urol, 59, pp.166~176 (2018)
- Gravas S, "Summary paper on the 2023 European Association of Urology guidelines on the management of non-neurogenic male lower urinary tract symptoms", Eur Urol, 84, pp.207~222 (2023)
- Cornu JN, "A systematic review and meta-analysis of functional outcomes and complications following transurethral procedures for lower urinary tract symptoms resulting from benign prostatic obstruction : an update", Eur Urol, 67, pp.1066~1096 (2015)
- Baboudjian M, "Summary paper on underactive bladder from the European Association of Urology guidelines on non-neurogenic male lower urinary tract symptoms", Eur Urol, 86, pp.213~220 (2024)
- Thomas AW, "The natural history of lower urinary tract dysfunction in men : the influence of detrusor underactivity on the outcome after transurethral resection of the prostate with a minimum 10-year urodynamic follow-up", BJU Int, 93, pp.745~750 (2004)
- Han DH, "The efficacy of transurethral resection of the prostate in the patients with weak bladder contractility index", Urology, 71, pp.657~661 (2008)
- Lomas DJ, "Long-term efficacy of holmium laser enucleation of the prostate in patients with detrusor underactivity or acontractility", Urology, 97, pp.208~211 (2016)
- Dobberfuhl AD, "Spontaneous voiding is surprisingly recoverable via outlet procedure in men with underactive bladder and documented detrusor underactivity on urodynamics", Neurourol Urodyn, 38, pp.2224~2232 (2019)
- Wu SY, "Predictive factors for recovery of voiding function after transurethral prostate surgery in men with small prostate volume and very low detrusor contractility", Low Urin Tract Symptoms, 12, pp.41~46 (2020)
- LeCun Y, "Deep learning", Nature, 521, pp.436~444 (2015)
- Vapnik VN, "An overview of statistical learning theory", IEEE Trans Neural Netw, 10, pp.988~999 (1999)
- Schmidhuber J, "Deep learning in neural networks : an overview", Neural Netw, 61, pp.85~117 (2015)
- Breiman L, "Random forests", Mach Learn, 45, pp.5~32 (2001)
- Cortes C, "Support-vector networks", Mach Learn, 20, pp.273~297 (1995)
- He H, "Learning from imbalanced data", IEEE Trans Knowl Data Eng, 21, pp.1263~1284 (2009)
- Micci-Barreca D, "A preprocessing scheme for high-cardinality categorical attributes in classification and prediction problems", SIGKDD Explor Newsl, 3, pp.27~32 (2001)
- Stigler SM, "Gauss and the invention of least squares", Ann Stat, 9, pp.465~474 (1981)
- Altmann A, "Permutation importance : a corrected feature importance measure", Bioinformatics, 26, pp.1340~1347 (2010)
- Uren AD, "Definition and symptoms of underactive bladder", Investig Clin Urol, 58(Suppl 2), pp.S61~S67 (2017)
- Mitchell CR, "Efficacy of holmium laser enucleation of the prostate in patients with non-neurogenic impaired bladder contractility : results of a prospective trial", Urology, 83, pp.428~432 (2014)
- Kim M, "Effect of preoperative urodynamic detrusor underactivity on transurethral surgery for benign prostatic hyperplasia : a systematic review and meta-analysis", J Urol, 199, pp.237~244 (2018)
- Cho MC, "Effect of preoperative detrusor underactivity on long-term surgical outcomes of photovaporization and holmium laser enucleation in men with benign prostatic hyperplasia : a lesson from 5-year serial followup data", BJU Int, 123, pp.E34~E42 (2019)
- Ito H, "Predicting prostate surgery outcomes from standard clinical assessments of lower urinary tract symptoms to derive prognostic symptom and flowmetry criteria", Eur Urol Focus, 10, pp.197~204 (2024)