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: Oct 23, 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

Versatile anomaly detection method for medical images with semi-supervised flow-based generative models.

Hisaichi Shibata¹, Shouhei Hanaoka², Yukihiro Nomura³

¹Department of Computational Diagnostic Radiology and Preventive Medicine, The University of Tokyo Hospital, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan. sh@g.ecc.u-tokyo.ac.jp.

International Journal of Computer Assisted Radiology and Surgery

|August 25, 2021

Summary

Related Concept Videos

Magnetic Resonance Imaging

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is a noninvasive medical imaging technique based on a phenomenon of nuclear physics discovered in the 1930s, in which matter exposed to magnetic fields and radio waves was found to emit radio signals. In 1970, a physician and researcher named Raymond Damadian noticed that malignant (cancerous) tissue gave off different signals than normal body tissue. He applied for a patent for the first MRI scanning device in clinical use by the early 1980s. The early MRI...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

A U-turn wiring technique for reverse-angled branches adjacent to ectasia: a case series.

European heart journal. Case reports·2024

Same author

Endoscopic Ventriculocisternostomy with Stent Placement for Trapped Temporal Horn.

World neurosurgery·2024

Same author

Bifurcation or periphery: Where should aberrant arteries in pulmonary sequestration be divided?

International journal of surgery case reports·2024

Same author

Scleroderma myocarditis with severe secondary mitral regurgitation successfully treated with transcatheter edge-to-edge repair: a case report.

European heart journal. Case reports·2024

Same author

Change-of-Direction Performance and Its Deficits in Relation to Countermovement-Jump Height and Phase-Specific Performance Among Female Athletes.

International journal of sports physiology and performance·2024

Same author

A periodic split attractor reconstruction method facilitates cardiovascular signal diagnoses and obstructive sleep apnea syndrome monitoring.

Heliyon·2024

This summary is machine-generated.

This study introduces a novel method using two flow-based deep generative (FDG) models to predict medical image normality. The approach effectively detects abnormalities in chest X-rays and brain CT scans, aiding radiologists.

Area of Science:

Medical imaging analysis
Deep learning in radiology
Bayesian inference in healthcare

Background:

Radiologists face heavy workloads due to the high volume of medical image interpretation.
Timely interpretation of medical images is crucial for effective clinical practice.
Developing automated methods can help alleviate radiologist workload and improve efficiency.

Purpose of the Study:

To formulate and validate a novel method for detecting virtually all types of lesions in medical images.
To reduce the workload of radiologists by providing a tool for efficient image interpretation.
To demonstrate the capability of two flow-based deep generative (FDG) models in predicting logarithm posterior probability for medical images in a semi-supervised manner.

Main Methods:

Keywords:

Anomaly detection Brain computed tomography Chest X-ray Deep learning Semi-supervised

Related Experiment Videos

Last Updated: Oct 23, 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

Utilized two flow-based deep generative (FDG) models in conjunction with Bayes' theorem.

Predicted the logarithm posterior probability of a medical image being normal.

Trained one FDG model exclusively on normal images and the second on both normal and non-normal images.

Main Results:

Validated the method on chest X-ray images (CXRs) and brain computed tomography (CT) images.
Achieved an average area under the receiver operating characteristic curve of 0.839 for detecting pneumonia-like opacities in CXRs.
Demonstrated an average area under the receiver operating characteristic curve of 0.904 for detecting infarction in BCTs.

Conclusions:

Successfully formulated a method using two FDG models to predict logarithm posterior probability.
Validated the method's ability to detect abnormal findings in CXRs and BCTs with acceptable performance.
The proposed method offers a comparatively light training workload while maintaining effective detection capabilities.