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 24, 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

Medical Referring Image Segmentation via Next-Token Mask Prediction.

Xinyu Chen, Yiran Wang, Gaoyang Pang

IEEE Transactions on Medical Imaging

|June 22, 2026

Summary

This summary is machine-generated.

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

Prototype-Based Multi-Dimension Intensity Mapping Density Sampling Network for Corrosion Segmentation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same author

OnUVS: An Online Motion Transfer Framework with Content-Texture Decoupling for High-Fidelity Ultrasound Video Synthesis.

IEEE journal of biomedical and health informatics·2026

Same author

A novel width-expanding spiral microchannel for high-throughput continuous separation of bacterial-sized particles.

Scientific reports·2026

Same author

Dean migration mechanisms: facilitating the control of microbial contaminants in cell therapy product manufacturing.

Microsystems & nanoengineering·2026

Same author

Universal modulus-free transfer of scalable laser-induced graphene for electronic skins.

Nature communications·2026

Same author

Comparative effectiveness of Hounsfield unit, MRI-based bone quality, dual-energy X-ray absorptiometry T -score, and quantitative computed tomography-measured bone mineral density in predicting cage subsidence and screw loosening post-spinal surgery.

International journal of surgery (London, England)·2026

Same journal

Informed-Exploration Reinforcement Learning for Automated Virtual Coronary Intervention Planning.

IEEE transactions on medical imaging·2026

Same journal

4D Reconstruction of Fetal Left Ventricle from Echocardiography via 2.5D Radial Segmentation and Graph-Fourier Reconstruction.

IEEE transactions on medical imaging·2026

Same journal

Generalised Medical Phrase Grounding.

IEEE transactions on medical imaging·2026

Same journal

EndoLRMGS: Combining Large Reconstruction Modelling and Gaussian Splatting for Complete Endoscopic Scene Reconstruction.

IEEE transactions on medical imaging·2026

Same journal

A Neural-Analytical Fusion Scatter Correction Method for Multi-Source CT Using Equivalent High-Order Scatter.

IEEE transactions on medical imaging·2026

Same journal

SynReEM: Synapse Reconstruction via Instance Structure Encoding in Anisotropic Electron Microscopic Volumes.

IEEE transactions on medical imaging·2026

See all related articles

NTP-MRISeg reformulates medical referring image segmentation as next-token prediction, simplifying model design and improving performance. This novel approach enhances generalization and adaptability for medical image analysis.

Area of Science:

Artificial Intelligence
Medical Imaging
Computer Vision

Background:

Medical Referring Image Segmentation (MRIS) typically requires complex multimodal fusion or multi-stage decoders.
Existing methods face challenges in streamlining model design and achieving end-to-end training.

Purpose of the Study:

To propose NTP-MRISeg, a novel framework that reformulates MRIS as an autoregressive next-token prediction task.
To streamline MRIS model design by eliminating the need for modality-specific fusion and external segmentation models.
To enhance generalization and adaptability by leveraging pretrained tokenizers from large-scale multimodal models.

Main Methods:

Unified multimodal sequence of tokenized image, text, and mask representations for autoregressive prediction.

More Related Videos

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

Related Experiment Videos

Last Updated: Jun 24, 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

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

Next-k Token Prediction (NkTP) scheme to mitigate cumulative prediction errors.

Token-level Contrastive Learning (TCL) for improved boundary sensitivity and long-tail distribution handling.

Memory-based Hard Error Token (HET) optimization to focus on difficult tokens.

Main Results:

NTP-MRISeg achieves new state-of-the-art performance on QaTa-COV19 and MosMedData+ datasets.
Demonstrates a streamlined and effective alternative to traditional MRIS pipelines.
Validates the effectiveness of NkTP, TCL, and HET strategies in addressing formulation challenges.

Conclusions:

NTP-MRISeg offers a simplified yet powerful framework for MRIS.
The autoregressive next-token prediction approach enhances model design and training efficiency.
The proposed strategies effectively address key challenges in the MRIS formulation, leading to superior performance.