Dataset Information

An automated approach for predicting glioma grade and survival of LGG patients using CNN and radiomics.

ABSTRACT:

Objectives

To develop and validate an efficient and automatically computational approach for stratifying glioma grades and predicting survival of lower-grade glioma (LGG) patients using an integration of state-of-the-art convolutional neural network (CNN) and radiomics.

Method

This retrospective study reviewed 470 preoperative MR images of glioma from BraTs public dataset (n=269) and Jinling hospital (n=201). A fully automated pipeline incorporating tumor segmentation and grading was developed, which can avoid variability and subjectivity of manual segmentations. First, an integrated approach by fusing CNN features and radiomics features was employed to stratify glioma grades. Then, a deep-radiomics signature based on the integrated approach for predicting survival of LGG patients was developed and subsequently validated in an independent cohort.

Results

The performance of tumor segmentation achieved a Dice coefficient of 0.81. The intraclass correlation coefficients (ICCs) of the radiomics features between the segmentation network and physicians were all over 0.75. The performance of glioma grading based on integrated approach achieved the area under the curve (AUC) of 0.958, showing the effectiveness of the integrated approach. The multivariable Cox regression results demonstrated that the deep-radiomics signature remained an independent prognostic factor and the integrated nomogram showed significantly better performance than the clinical nomogram in predicting overall survival of LGG patients (C-index: 0.865 vs. 0.796, P=0.005).

Conclusion

The proposed integrated approach can be noninvasively and efficiently applied in prediction of gliomas grade and survival. Moreover, our fully automated pipeline successfully achieved computerized segmentation instead of manual segmentation, which shows the potential to be a reproducible approach in clinical practice.

SUBMITTER: Xu C

PROVIDER: S-EPMC9413530 | biostudies-literature | 2022

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

An automated approach for predicting glioma grade and survival of LGG patients using CNN and radiomics.

Xu Chenan C Peng Yuanyuan Y Zhu Weifang W Chen Zhongyue Z Li Jianrui J Tan Wenhao W Zhang Zhiqiang Z Chen Xinjian X

Frontiers in oncology 20220812

<h4>Objectives</h4>To develop and validate an efficient and automatically computational approach for stratifying glioma grades and predicting survival of lower-grade glioma (LGG) patients using an integration of state-of-the-art convolutional neural network (CNN) and radiomics.<h4>Method</h4>This retrospective study reviewed 470 preoperative MR images of glioma from BraTs public dataset (n=269) and Jinling hospital (n=201). A fully automated pipeline incorporating tumor segmentation and grading ...[more]

PMID: 36033433

Similar Datasets

Project description:BackgroundPyroptosis, a lytic form of programmed cell death initiated by inflammasomes, has been reported to be closely associated with tumor proliferation, invasion and metastasis. However, the roles of pyroptosis genes (PGs) in low-grade glioma (LGG) remain unclear.MethodsWe obtained information for 1,681 samples, including the mRNA expression profiles of LGGs and normal brain tissues and the relevant corresponding clinical information from two public datasets, TCGA and GTEx, and identified 45 differentially expressed pyroptosis genes (DEPGs). Among these DEPGs, nine hub pyroptosis genes (HPGs) were identified and used to construct a genetic risk scoring model. A total of 476 patients, selected as the training group, were divided into low-risk and high-risk groups according to the risk score. The area under the curve (AUC) values of the receiver operating characteristic (ROC) curves verified the accuracy of the model, and a nomogram combining the risk score and clinicopathological characteristics was used to predict the overall survival (OS) of LGG patients. In addition, a cohort from the Gene Expression Omnibus (GEO) database was selected as a validation group to verify the stability of the model. qRT-PCR was used to analyze the gene expression levels of nine HPGs in paracancerous and tumor tissues from 10 LGG patients.ResultsSurvival analysis showed that, compared with patients in the low-risk group, patients in the high-risk group had a poorer prognosis. A risk score model combining PG expression levels with clinical features was considered an independent risk factor. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses indicated that immune-related genes were enriched among the DEPGs and that immune activity was increased in the high-risk group.ConclusionIn summary, we successfully constructed a model to predict the prognosis of LGG patients, which will help to promote individualized treatment and provide potential new targets for immunotherapy.

Project description:IntroductionGlioma is the most common primary tumor in the brain.Integrin beta 2(ITGB2) is a member of the leukocyte integrin family (leukocyte integrin), participating in lymphocyte recycling and homing, cell adhesion, and cell surface-mediated signal transduction. However, few studies on ITGB2 in gliomas have been reported yet.This study first discussed the relationship between ITGB2 expression and clinical characterization of glioma and the prognostic significance of its methylation in low-grade glioma.MethodsWe collected Clinical data and transcription of glioma patients from TCGA, CGGA, and Rembrant datasets to analyze the differential expression of ITGB2 mRNA in glioma tissues and normal tissues. The box polts to evaluated the expression patterns of ITGB2 in different molecular subtypes. Receiver operating characteristic curve (ROC) were used to evaluate and verify the reliability of the model. Kaplan-Meier survival curves to evaluated the relationship between the level of ITGB2 mRNA expression and overall survival (OS). Using cox regression analysis to verify the ability of ITGB2 as an independent predictor of OS in glioma patients. We use TIMER to analyze and visualize the association between immune infiltration levels and a range of variables. The methylation of GBMLGG patients were obtained from the TCGA database through the biological portal.ResultsITGB2 can be a potential marker for mesenchymal molecular subtype gliomas. COX regression analysis shows that ITGB2 is an independent predictive marker of OS in malignant glioma patients. Biological processes show that ITGB2 has involved glioma immune-related activities, especially closely related to B cells, CD4+Tcells, macrophages, neutrophils, and dendritic cells. ITGB2 is negatively regulated by ITGB2 methylation, resulting in low expression in LGG tissues. Low expression of ITGB2 and high methylation indicate good OS in patients with LGG. The ITGB2 methylation risk score (ITMRS) obtained from the ITGB2 methylation CpG site can better predict the OS of LGG patients. We used univariate and multivariate cox regression analysis of methylationsites, used the R language predict function to obtain the risk score of these ITGB2 methylation sites(ITMRS).DiscussionITGB2 can be used as a potential marker of mesenchymal molecular subtypes of gliomas and as an independent predictive marker of OS in patients with malignant gliomas. The ITMRS we established can be used as an independent prognostic factor for LGG and provide a new idea for the diagnosis and treatment of LGG.

Dataset Information

An automated approach for predicting glioma grade and survival of LGG patients using CNN and radiomics.

Objectives

Method

Results

Conclusion

Publications

An automated approach for predicting glioma grade and survival of LGG patients using CNN and radiomics.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets