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

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

Two-Staged Versus Three-Staged Laparoscopic Anorectoplasty for Patients with Rectoprostatic and Bladder Neck Fistulas: A Comparative Study.

Journal of laparoendoscopic & advanced surgical techniques. Part A·2019

Same author

Genome Sequence of a Novel HIV-1 Circulating Recombinant Form (CRF103_01B) Identified from Hebei Province, China.

AIDS research and human retroviruses·2019

Same author

Identification of a Novel HIV-1 Second-Generation (CRF01_AE/B) Among Men Who Have Sex with Men in Tianjin, China.

AIDS research and human retroviruses·2019

Same author

Characterization of a Novel HIV-1 Recombinant Form (CRF01_AE/CRF07_BC/CRF08_BC) Identified from Guangxi, China.

AIDS research and human retroviruses·2019

Same author

Event-Triggered Multiagent Optimization for Two-Layered Model of Hybrid Energy System With Price Bidding-Based Demand Response.

IEEE transactions on cybernetics·2019

Same author

A Human Peripheral Blood Mononuclear Cell (PBMC) Engrafted Humanized Xenograft Model for Translational Immuno-oncology (I-O) Research.

Journal of visualized experiments : JoVE·2019

Search research articles

Related Experiment Video

Updated: Aug 1, 2025

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

Self-supervised anomaly detection, staging and segmentation for retinal images.

Yiyue Li¹, Qicheng Lao², Qingbo Kang³

¹Department of Ophthalmology and West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, Sichuan, 610041, China; West China Biomedical Big Data Center, Med-X Center for Informatics, Sichuan University, Chengdu, Sichuan, 610041, China.

Medical Image Analysis

|April 27, 2023

Summary

This summary is machine-generated.

This study introduces SSL-AnoVAE, a novel framework for unsupervised anomaly detection in retinal images. It enhances anomaly detection, staging, and segmentation using self-supervised learning for improved computer-aided diagnosis.

Keywords:

Anomaly detection Anomaly segmentation Anomaly staging Retinal images

More Related Videos

Using Retinal Imaging to Study Dementia

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

Efficient and Consistent Generation of Retinal Pigment Epithelium/Choroid Flatmounts from Human Eyes for Histological Analysis

Efficient and Consistent Generation of Retinal Pigment Epithelium/Choroid Flatmounts from Human Eyes for Histological Analysis

Published on: October 28, 2022

Related Experiment Videos

Last Updated: Aug 1, 2025

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

Using Retinal Imaging to Study Dementia

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

Efficient and Consistent Generation of Retinal Pigment Epithelium/Choroid Flatmounts from Human Eyes for Histological Analysis

Efficient and Consistent Generation of Retinal Pigment Epithelium/Choroid Flatmounts from Human Eyes for Histological Analysis

Published on: October 28, 2022

Area of Science:

Medical Imaging
Computer Vision
Artificial Intelligence

Background:

Unsupervised anomaly detection (UAD) in medical images aids diagnosis by learning normal data distributions without labels.
Existing UAD methods often neglect semantic prior information, limiting their effectiveness.
Retinal image analysis benefits from UAD due to the high cost of annotated data.

Purpose of the Study:

To propose a universal unsupervised anomaly detection framework (SSL-AnoVAE) for retinal images.
To integrate self-supervised learning (SSL) for enhanced semantic understanding of anomalies.
To enable unsupervised anomaly staging and segmentation for clinical applications in retinal diseases.

Main Methods:

Developed the SSL-AnoVAE framework incorporating an SSL module for fine-grained semantic anomaly detection.
Investigated the impact of data transformations within the SSL module on anomaly detection performance.
Extended the framework for unsupervised anomaly staging and segmentation in retinal images.

Main Results:

SSL-AnoVAE demonstrated effectiveness in unsupervised anomaly detection, staging, and segmentation.
The study explored the relationship between SSL data transformations and anomaly detection quality.
The proposed method showed significant results on both optical coherence tomography and fundus photograph images.

Conclusions:

The SSL-AnoVAE framework offers a robust solution for unsupervised anomaly detection in retinal imaging.
The integration of SSL significantly improves the semantic understanding of anomalies.
The method holds promise for computer-aided diagnosis of retinal diseases, facilitating unsupervised staging and segmentation.