Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

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

Low-frequency bipolar versus high-frequency monopolar cortical and subcortical motor mapping in craniotomies for brain tumors.

Journal of neurosurgery·2026
Same author

Deep learning-derived pericardial adipose tissue by electrocardiogram-gated cardiac computed tomography predicts cardiovascular events beyond coronary calcium score.

American journal of preventive cardiology·2026
Same author

A call for action: The need to quantify the "-itis" in primary sclerosing cholangitis.

Journal of hepatology·2026
Same author

Evaluating the current neuro-oncology capacity in Sub-Saharan Africa: A questionnaire-based survey.

Neuro-oncology practice·2026
Same author

Impact of modern central nervous system radiation on multiple sclerosis relapse and progression.

Clinical and translational radiation oncology·2026
Same author

MRI Deep Learning for Differentiating Glioblastoma, IDH Wild-type from Central Nervous System Diffuse Large B-cell Lymphoma.

Cancer research communications·2026

Related Experiment Video

Updated: Feb 25, 2026

Comprehensive DNA Methylation Analysis Using a Methyl-CpG-binding Domain Capture-based Method in Chronic Lymphocytic Leukemia Patients
13:21

Comprehensive DNA Methylation Analysis Using a Methyl-CpG-binding Domain Capture-based Method in Chronic Lymphocytic Leukemia Patients

Published on: June 16, 2017

10.6K

Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status.

Panagiotis Korfiatis1, Timothy L Kline1, Daniel H Lachance2

  • 1Department of Radiology, Mayo Clinic, 200 1st Street SW, Rochester, MN, 55905, USA.

Journal of Digital Imaging
|August 9, 2017
PubMed
Summary

Deep learning models can predict O6-methylguanine methyltransferase (MGMT) gene methylation status from MRI scans. ResNet50 achieved 94.90% accuracy, outperforming other ResNet models for brain tumor prognosis.

Keywords:
Deep learningMGMT methylationMRI

More Related Videos

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging
11:38

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

1.1K
Sample Preparation to Bioinformatics Analysis of DNA Methylation: Association Strategy for Obesity and Related Trait Studies
14:56

Sample Preparation to Bioinformatics Analysis of DNA Methylation: Association Strategy for Obesity and Related Trait Studies

Published on: May 6, 2022

5.2K

Related Experiment Videos

Last Updated: Feb 25, 2026

Comprehensive DNA Methylation Analysis Using a Methyl-CpG-binding Domain Capture-based Method in Chronic Lymphocytic Leukemia Patients
13:21

Comprehensive DNA Methylation Analysis Using a Methyl-CpG-binding Domain Capture-based Method in Chronic Lymphocytic Leukemia Patients

Published on: June 16, 2017

10.6K
Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging
11:38

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

1.1K
Sample Preparation to Bioinformatics Analysis of DNA Methylation: Association Strategy for Obesity and Related Trait Studies
14:56

Sample Preparation to Bioinformatics Analysis of DNA Methylation: Association Strategy for Obesity and Related Trait Studies

Published on: May 6, 2022

5.2K

Area of Science:

  • Neuro-oncology
  • Medical Imaging
  • Artificial Intelligence

Background:

  • O6-methylguanine methyltransferase (MGMT) gene methylation status is crucial for predicting treatment response and prognosis in brain tumors.
  • Current prediction methods often require invasive procedures or complex analyses.
  • Non-invasive prediction using routine medical imaging is highly desirable.

Purpose of the Study:

  • To evaluate the efficacy of different residual deep neural network (ResNet) architectures in predicting MGMT methylation status directly from MRI images.
  • To determine if deep learning can bypass the need for explicit tumor segmentation in this prediction task.

Main Methods:

  • Comparison of three ResNet architectures (ResNet18, ResNet34, ResNet50) for classifying MRI slices.
  • Classification categories included: no tumor, methylated MGMT, and non-methylated MGMT.
  • Statistical analysis to compare the performance of different models.

Main Results:

  • ResNet50 achieved the highest accuracy (94.90% ± 3.92%) in predicting MGMT methylation status.
  • ResNet50 performance was statistically significantly superior to ResNet18 (76.75% ± 20.67%) and ResNet34 (80.72% ± 13.61%) (p < 0.001).
  • The developed method requires minimal preprocessing, demonstrating a proof of concept for deep learning in biomarker prediction.

Conclusions:

  • Deep neural networks, particularly ResNet50, can accurately predict MGMT methylation status from routine MRI scans.
  • This approach offers a non-invasive method for predicting a key molecular biomarker, potentially improving brain tumor management.
  • The study highlights the potential of AI in medical imaging for non-invasive molecular biomarker discovery.