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: Jun 29, 2026

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

Deployable real-time spinal endoscopic instance segmentation with lightweight multi-scale attention mechanism.

Qi Lai¹, Qiang Cai², JunYan Li³

¹Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, PR China.

NPJ Digital Medicine

|June 10, 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

Galangin Ameliorates Acetaminophen-Induced Acute Liver Injury by Inhibiting Ferroptosis through the Prevention of Chaperone-Mediated Autophagic Degradation of FTH1.

Journal of agricultural and food chemistry·2026

Same author

Dibutyl phthalate induces sarcopenia via TNFα/TNFR1-mediated proteolytic and pyroptotic axes: evidence from NHANES and experimental models.

Frontiers in immunology·2026

Same author

Construction of Nitrogen-rich Graphene Oxide/Chitosan Adsorbent for the Removal of Hexavalent Chromium from Wastewater.

Journal of separation science·2026

Same author

Beyond wound closure: translational opportunities and barriers of mesenchymal stem cells and their extracellular vesicles in burn management.

Frontiers in cell and developmental biology·2026

Same author

Advances in precision synthesis and synergistic applications of Janus nanostructures.

Advances in colloid and interface science·2026

Same author

Corrigendum to "Collaborate Large and Small Language Models for Multi-Modal Emergency Rumor Detection" [Neural Networks 190, 2025, 107625].

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Explainable foundation model for dementia screening and risk stratification using retinal fundus images.

NPJ digital medicine·2026

Same journal

LLM research on public biosignals data is needed to protect patients.

NPJ digital medicine·2026

Same journal

Conversational artificial intelligence for pre-procedural patient preparation: implementation, validation and patient satisfaction.

NPJ digital medicine·2026

Same journal

Whole body CT attenuation and volume charts from routine clinical scans via LLM report filtering.

NPJ digital medicine·2026

Same journal

Fast information and slow evidence in the large language models era.

NPJ digital medicine·2026

Same journal

Predicting response to neoadjuvant therapy using artificial intelligence on digitized histopathology slides: a systematic review.

NPJ digital medicine·2026

See all related articles

This summary is machine-generated.

EndoSeg-RT offers real-time instance segmentation for spinal endoscopy, improving anatomical identification despite challenging conditions. This lightweight framework achieves high accuracy and speed, even with single-image processing.

Area of Science:

Medical Imaging
Computer Vision
Surgical Technology

Background:

Real-time instance segmentation in spinal endoscopy is crucial for identifying and protecting critical anatomy.
Challenges include narrow views, artifacts (specular highlights, smoke, bleeding), fuzzy boundaries, and scale variations.
Deployment requires accuracy, speed, and stability, especially in small-batch (batch-size-one) settings.

Purpose of the Study:

To present EndoSeg-RT, a deployable real-time framework for spinal endoscopic instance segmentation.
To address the limitations of existing methods in challenging surgical environments.
To provide a lightweight yet accurate solution for real-time anatomical identification.

Main Methods:

Developed EndoSeg-RT, a framework with a lightweight multi-scale attention mechanism across backbone, neck, and head.

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Related Experiment Videos

Last Updated: Jun 29, 2026

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Introduced C2f-Pro backbone combining re-parameterized convolutions and efficient multi-scale attention.

Implemented Scale-Sequence Feature Fusion and Triple Feature Encoding in the neck for enhanced feature consistency and boundary quality.

Utilized a Lightweight Multi-task Shared Head with shared convolutions and GroupNorm for efficiency and stability.

Main Results:

EndoSeg-RT achieved competitive or superior accuracy compared to heavier instance segmentation baselines.
The framework boasts significantly lower complexity with only 1.8M parameters and 8.8 GFLOPs.
Demonstrated strong generalization capabilities on a public dental instance segmentation benchmark.
Released the PELD dataset with clinical review and instance-level labels for key spinal structures.

Conclusions:

EndoSeg-RT provides an effective, efficient, and deployable solution for real-time instance segmentation in spinal endoscopy.
The framework's lightweight design and robust performance address critical needs in surgical visualization.
The publicly available code and dataset facilitate further research and development in medical image analysis.