Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Apr 16, 2026

Retzius-Sparing Robot-Assisted Radical Prostatectomy

Retzius-Sparing Robot-Assisted Radical Prostatectomy

Published on: May 19, 2022

Deep-learning segmentation to guide bladder neck recognition in robot-assisted radical prostatectomy.

Yoshinari Muto¹, Kenji Zennami^1,2, Kota Yagi^1,3

¹Department of Urology, Fujita Health University School of Medicine, Toyoake, Japan.

BJU International

|April 15, 2026

Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

<i>ASPSCR1: TFE3</i> Fusion-Positive Epithelioid Angiomyolipoma/PEComa Lacking Melanocytic Markers: A Case Report.

IJU case reports·2026

Same author

Feasibility and Early Outcomes of Single-Port Robot-Assisted Partial Nephrectomy via Supine Low Anterior Access.

Asian journal of endoscopic surgery·2026

Same author

Incidence and Risk Factors for Perioperative Infections After Robot-Assisted Radical Prostatectomy: A Multicenter Study.

International journal of urology : official journal of the Japanese Urological Association·2026

Same author

Development of a prognostic model for elderly patients with metastatic clear cell renal cell carcinoma treated with first-line immuno-oncology combination therapy.

Japanese journal of clinical oncology·2026

Same author

Author's reply to 'Bayesian interpretation of febrile neutropenia risk and primary G-CSF prophylaxis during darolutamide triplet therapy'.

Japanese journal of clinical oncology·2026

Same author

Genomic Profiling in Localized Prostate Cancer: Associations With Biochemical Recurrence and Response to Salvage Radiotherapy.

Cancer science·2026

A deep learning model for bladder neck dissection in robot-assisted radical prostatectomy significantly improved anatomical recognition for novice surgeons. This AI tool enhances surgical video analysis and may support training in urology.

Area of Science:

Medical image analysis
Artificial intelligence in surgery
Urological oncology

Background:

Robot-assisted radical prostatectomy (RARP) is a complex procedure requiring precise anatomical identification.
Bladder neck dissection is a critical step in RARP, with anatomical recognition challenges, especially for less experienced surgeons.
Current methods for anatomical recognition in surgical videos lack automated assistance.

Purpose of the Study:

To develop and evaluate a deep-learning segmentation model for bladder neck dissection during RARP.
To assess the model's effectiveness in improving anatomical recognition across surgeons with varying experience levels.
To investigate the clinical utility of the AI model in enhancing surgical video analysis.

Main Methods:

Keywords:

active learning bladder neck dissection deep learning education robot‐assisted radical prostatectomy segmentation

More Related Videos

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Related Experiment Videos

Last Updated: Apr 16, 2026

Retzius-Sparing Robot-Assisted Radical Prostatectomy

Retzius-Sparing Robot-Assisted Radical Prostatectomy

Published on: May 19, 2022

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Trained three deep learning architectures (U-Net, U-Net++, DeepLabv3+) on annotated RARP surgical video frames.

Employed a two-step, entropy-based active learning strategy to enhance model generalisability.

Evaluated segmentation performance using mean Intersection over Union (mIoU) and clinical utility by measuring annotation time and bladder-prostate IoU (bpIoU) among novice and expert urologists.

Main Results:

DeepLabv3+ achieved the highest segmentation performance (mIoU = 0.815).
Active learning improved model generalisability, with a slight increase in mIoU on the novice evaluation set.
For novice surgeons, the model significantly reduced annotation time by 28 seconds and improved bpIoU by 0.124, with no significant impact on expert surgeons.

Conclusions:

The developed deep-learning segmentation model effectively enhances anatomical recognition speed and accuracy, particularly for novice surgeons.
The AI tool shows potential for supporting surgical education and improving procedural efficiency in RARP.
Automated segmentation in surgical videos can bridge the experience gap in complex procedures.