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

Unexpected detection of clear cell sarcoma of soft tissue during single-channel endoscopic carpal tunnel release for recurrent carpal tunnel syndrome: a case report with literature review.

Frontiers in surgery·2026

Same author

Harnessing the Janus-faced nature of polyoxometalates to regulate cellular redox homeostasis.

Chemical communications (Cambridge, England)·2026

Same author

Platelet gene signatures detecting pulmonary artery stenosis in patients with pulmonary hypertension.

Orphanet journal of rare diseases·2026

Same author

ColoLDB: a machine learning-based predictive model for colorectal cancer using routine laboratory parameters.

Journal of gastrointestinal oncology·2026

Same author

Development and Evaluation of Artificial Intelligence-Based Two-Step Model for Automated Serum Quality Assessment in Clinical Laboratories.

Annals of laboratory medicine·2026

Same author

Association of VitD 3 deficiency with thyroid nodules suspected of malignancy in petroleum workers: a retrospective cohort study.

PeerJ·2026

Search research articles

Related Experiment Video

Updated: Jun 1, 2025

Digital Spatial Profiling for Characterization of the Microenvironment in Adult-Type Diffusely Infiltrating Glioma

Digital Spatial Profiling for Characterization of the Microenvironment in Adult-Type Diffusely Infiltrating Glioma

Published on: September 13, 2022

Comparative analysis of deep learning and radiomic signatures for overall survival prediction in recurrent high-grade

Qi Wan^1,2, Clifford Lindsay³, Chenxi Zhang⁴

¹Department of Radiology, the Key Laboratory of Advanced Interdisciplinary Studies Center, the First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China. qiwan@gzhmu.edu.cn.

Cancer Imaging : the Official Publication of the International Cancer Imaging Society

|January 22, 2025

Summary

Manual segmentation radiomic features better predict survival in glioma patients than automated methods. An end-to-end deep learning model showed similar performance without segmentation, offering a potential time-saving alternative.

Keywords:

Convolutional neural networks Deep learning High-grade glioma Overall survival Radiomics

More Related Videos

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Predictive Immune Modeling of Solid Tumors

Predictive Immune Modeling of Solid Tumors

Published on: February 25, 2020

Related Experiment Videos

Last Updated: Jun 1, 2025

Digital Spatial Profiling for Characterization of the Microenvironment in Adult-Type Diffusely Infiltrating Glioma

Digital Spatial Profiling for Characterization of the Microenvironment in Adult-Type Diffusely Infiltrating Glioma

Published on: September 13, 2022

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Predictive Immune Modeling of Solid Tumors

Predictive Immune Modeling of Solid Tumors

Published on: February 25, 2020

Area of Science:

Radiology and Medical Imaging
Oncology
Artificial Intelligence in Medicine

Background:

Radiomic analysis aids prognosis prediction in brain tumor patients.
Manual tumor segmentation is time-consuming and lacks reproducibility.
This study compares manual and automated segmentation for predicting overall survival (OS) in recurrent high-grade glioma (HGG) patients undergoing immunotherapy.

Purpose of the Study:

To compare the predictive performance of radiomic features derived from manual versus automated (CNN) segmentation for overall survival (OS) in recurrent high-grade glioma (HGG) patients.
To evaluate an end-to-end deep learning (DL) model for OS prediction in the same patient cohort.
To assess the impact of robust feature selection on prediction accuracy.

Main Methods:

Retrospective analysis of 154 recurrent HGG cases.
Tumor segmentation using expert radiologists and a convolutional neural network (CNN).
Extraction and selection of 2553 radiomic features, with robust subset selection.
Support Vector Machine (SVM) for radiomics-based OS prediction and CNN for direct classification.
Ten-fold cross-validation and rotating test set for model validation.

Main Results:

Manual segmentation-derived radiomics achieved higher OS prediction accuracy (AUC 0.662) than automated segmentation (AUC 0.471) in the test set.
Robust feature selection improved manual segmentation performance (AUC 0.700) and automated segmentation performance (AUC 0.554).
The end-to-end CNN prognosis model demonstrated comparable performance to manual radiomics (AUC 0.700).

Conclusions:

Radiomic features from manual segmentation are superior for predicting OS in HGG patients undergoing immunotherapy.
An end-to-end CNN model offers comparable predictive performance to manual radiomics without requiring segmentation.
The trade-off between time-saving automated methods and the interpretability of manual segmentation requires consideration.