Deep learning based automated quantification of urethral plate characteristics using the plate objective scoring tool (POST)

IntroductionThe plate objective scoring tool (POST) was recently introduced as a reproducible and precise approach to quantifying urethral plate (UP) characteristics and guide to selecting particular surgical tech-niques. However, defining the landmarks mandatory for the POST score from captured images can potentially leads to variability. Although artificial intelligence (AI) is yet to be wholly accepted and explored in hypospadiology, it has certainly brought new possibilities to light.ObjectivesTo explore the capacity of deep learning algorithm to further streamline and optimize UP characteris-tics appraisal on 2D images using the POST, aiming to increase the objectivity and reproducibility of UP appraisal in hypospadias repair.MethodsThe five key POST landmarks were marked by spe-cialists in a 691-image dataset of prepubertal boys undergoing primary hypospadias repair. This dataset was then used to develop and validate a deep learning-based landmark detection model. The pro-posed framework begins with glans localization and detection, where the input image is cropped using the predicted bounding box. Next, a deep convolu-tional neural network (CNN) architecture is used to predict the coordinates of the five POST landmarks. These predicted landmarks are then used to assess UP characteristics in distal hypospadias.ResultsThe proposed model accurately localized the glans area, with a mean average precision (mAP) of 99.5% and an overall sensitivity of 99.1%. A normalized mean error (NME) of 0.07152 was achieved in pre-dicting the coordinates of the landmarks, with a mean squared error (MSE) of 0.001 and a 2.5% failure rate at a threshold of 0.2 NME.DiscussionOur results support the possibility of further stan-dardizing UP assessment from captured hypospadias images, and the use of machine learning algorithms and image recognition shows that these novel arti-ficial intelligence technologies are useful for scoring hypospadias. External validation can provide valu-able information on the generalizability and reli-ability of deep learning algorithms, which can aid in assessments, decision-making and predictions for surgical outcomes.ConclusionsThis deep learning application shows robustness and high precision in using POST to appraise UP charac-teristics. Further assessment using international multi-centre image-based databases is ongoing.

Automated summary

This study explores the use of deep learning algorithm to assess urethral plate characteristics in 2D images, aiming to increase the objectivity and reproducibility of hypospadias repair. A deep learning-based landmark detection model was developed and validated using a dataset of prepubertal boys. The model accurately localized the glans area and predicted the coordinates of the landmarks, leading to an accurate appraisal of UP characteristics in distal hypospadias.