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 Videos

Multi-View Stenosis Classification Leveraging Transformer-Based Multiple-Instance Learning Using Real-World Clinical

N Cenikj, O Turgut, A Muller

IEEE Transactions on Medical Imaging

|June 15, 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

Detection of microsporidia (Enterocytozoon bieneusi) in intestinal biopsy specimens from human immunodeficiency virus-infected patients by PCR.

Journal of clinical microbiology·1995

Same author

Parturition in the rabbit is compromised by daytime nursing: the role of oxytocin.

Biology of reproduction·1995

Same author

Isozyme selective inhibition of cGMP-stimulated cyclic nucleotide phosphodiesterases by erythro-9-(2-hydroxy-3-nonyl) adenine.

Cellular signalling·1995

Same author

[Residual placental blood and bone marrow as sources of hematopoietic stem cells for allogenic stem cell transplantation. Comparative analysis of hematopoietic potential].

Klinische Padiatrie·1995

Same author

Immediate metabolic effects of different nutritional regimens in critically ill medical patients.

Intensive care medicine·1995

Same author

Flybrain, an on-line atlas and database of the Drosophila nervous system.

Neuron·1995

Same journal

LLM-enhanced Neuron Segmentation and Reconstruction in Complex Mouse Brain Images.

IEEE transactions on medical imaging·2026

Same journal

Matrixed-Spectrum Decomposition Accelerated Linear Boltzmann Transport Equation Solver for Fast Scatter Correction in Multi-Spectral CT.

IEEE transactions on medical imaging·2026

Same journal

The Ritz Adjoint Method for MRI Pulse Design.

IEEE transactions on medical imaging·2026

Same journal

Physiology-guided Self-supervised Learning for Simultaneous Dual-Tracer PET Separation.

IEEE transactions on medical imaging·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

See all related articles

SegmentMIL, a new deep learning model, accurately diagnoses coronary artery stenosis using patient-level data without needing expensive view-specific labels. This approach improves cardiovascular disease diagnosis by analyzing multiple angiography views effectively.

Area of Science:

Cardiovascular Imaging and Diagnostics
Artificial Intelligence in Medicine
Medical Image Analysis

Background:

Coronary artery stenosis is a primary cause of cardiovascular disease.
Current deep learning models for stenosis detection require costly view-level annotations, limiting their clinical applicability.
Existing single-view models overlook crucial temporal dynamics and inter-view dependencies.

Purpose of the Study:

To develop a novel multi-view multiple-instance learning framework for patient-level coronary artery stenosis classification.
To address the limitations of view-level annotations in deep learning models for stenosis detection.
To create a clinically viable and scalable solution for diagnosing coronary artery stenosis.

Main Methods:

Proposed SegmentMIL, a transformer-based multi-view multiple-instance learning framework.

Related Experiment Videos

Utilized patient-level supervision on a real-world clinical dataset, eliminating the need for view-level annotations.

Jointly predicted stenosis presence and localized affected anatomical regions across coronary arteries and segments.

Main Results:

SegmentMIL achieved high performance in both internal and external evaluations.
The model outperformed existing view-level models and classical multiple-instance learning baselines.
Demonstrated accurate localization of stenosis in both right and left coronary arteries and their segments.

Conclusions:

SegmentMIL offers a scalable and clinically viable solution for coronary artery stenosis diagnosis.
The framework effectively utilizes patient-level supervision for multi-view analysis.
This approach overcomes the annotation bottleneck, enhancing diagnostic capabilities in real-world clinical settings.