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: Jul 15, 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

Uncertainty-guided mamba network for efficient medical image segmentation with evidential deep learning.

Siqin Sun¹, Chihui Long¹, Xingbo Dong²

¹Wuhan Third Hospital (Tongren Hospital of WuHan University), Wuhan, 430060, China.

Scientific Reports

|July 13, 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

Targeting SCAMP2 by a natural product auxarconjugatin B for glioblastoma therapy <i>via</i> restoring aspartate metabolic flux.

Acta pharmaceutica Sinica. B·2026

Same author

Hypoxia-induced endothelial exosomal MSTRG.12883.2 enhances ox-LDL- induced phenotypic switching of vascular smooth muscle cells.

Life sciences·2026

Same author

DeepFAN, a transformer-based model for human-artificial intelligence collaborative assessment of incidental pulmonary nodules in CT scans: a multireader, multicase trial.

Nature cancer·2026

Same author

An Interpretable Radiomics Model Based on Multi-DLCT Images for Differentiating Benign and Malignant Solid Solitary Pulmonary Nodules.

Medical physics·2026

Same author

Single-Cell Sequencing Data Revealed Mechanisms of Interactions Between Tumor Cells and Cancer-Associated Fibroblasts in Metastatic Colorectal Cancer.

BioMed research international·2026

Same author

Collaborative Coarse-to-Fine Disease Learning With Discharge Summary Awareness for EHR Event Prediction.

IEEE transactions on cybernetics·2026

This study introduces an efficient medical image segmentation framework using Mamba state-space models and evidential deep learning, achieving high accuracy with reduced computational cost and providing uncertainty quantification for clinical applications.

Area of Science:

Medical Image Analysis
Deep Learning
Computer Vision

Background:

Medical image segmentation demands a balance between accuracy, computational efficiency, and uncertainty quantification for clinical use.
Transformer models offer high accuracy but are computationally expensive, while CNNs are efficient but have limited receptive fields.

Purpose of the Study:

To develop an uncertainty-aware and computationally efficient medical image segmentation framework.
To leverage Mamba state-space models and evidential deep learning for improved segmentation performance and uncertainty estimation.

Main Methods:

Utilized a 2D-adapted selective state-space mechanism for efficient long-range dependency modeling with linear complexity.
Incorporated an uncertainty-guided attention module (UGAM) with Dirichlet-parameterized evidential learning to decompose and manage uncertainty.

Keywords:

Attention mechanism Clinical decision support Evidential deep learning Medical image segmentation Uncertainty quantification

Related Experiment Videos

Last Updated: Jul 15, 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

Employed progressive multi-scale fusion and gradient-based uncertainty supervision for enhanced boundary delineation.

Main Results:

Achieved competitive performance with 82.67% mean Dice on the Synapse dataset, outperforming Swin-UNet.
Demonstrated significant efficiency gains with only 7.8M parameters and 4.7 GFLOPs, enabling real-time inference at 37.2 FPS.
Showcased well-calibrated uncertainty quantification (Expected Calibration Error: 0.046).

Conclusions:

The proposed framework offers a promising balance of accuracy, efficiency, and uncertainty quantification for medical image segmentation.
Exploratory reader studies suggest potential benefits in reader confidence and decision time, warranting further validation.
The method is suitable for time-sensitive clinical workflow analysis, with future work addressing 3D volumes and other complex scenarios.