Radiomics in peritumoral non-enhancing regions: fractional anisotropy and cerebral blood volume improve prediction of local progression and overall survival in patients with glioblastoma

Kim, Jung Youn; Yoon, Min Jae; Park, Ji Eun; Choi, Eun Jung; Lee, Jong-Ho; Kim, Ho Sung

doi:10.1007/s00234-019-02255-4

Cited by 45 publications

(36 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hu et al, in 2015 used multi-parametric MRI and texture analysis to predict tumour density in both enhanced and non-enhanced part based on a biopsy study and showed 85% accuracy in training set and 81.8% accuracy in the validation 12 . Further study conducted by Kim et al, showed that the FA and the CBV radiomics of the peritumoural area, together with the clinical index, can potentially predict the tumour progression 13 .…”

mentioning

confidence: 99%

A Neural Network Approach to Identify the Peritumoral Invasive Areas in Glioblastoma Patients by Using MR Radiomics

Yan

Hoorn

et al. 2020

Sci Rep

View full text Add to dashboard Cite

the challenge in the treatment of glioblastoma is the failure to identify the cancer invasive area outside the contrast-enhancing tumour which leads to the high local progression rate. Our study aims to identify its progression from the preoperative MR radiomics. 57 newly diagnosed cerebral glioblastoma patients were included. All patients received 5-aminolevulinic acid (5-ALA) fluorescence guidance surgery and postoperative temozolomide concomitant chemoradiotherapy. Preoperative 3 T MRI data including structure MR, perfusion MR, and DTI were obtained. Voxel-based radiomics features extracted from 37 patients were used in the convolutional neural network to train and as internal validation. Another 20 patients of the cohort were tested blindly as external validation. Our results showed that the peritumoural progression areas had higher signal intensity in FLAIR (p = 0.02), rCBV (p = 0.038), and T1C (p = 0.0004), and lower intensity in ADC (p = 0.029) and DTI-p (p = 0.001) compared to non-progression area. The identification of the peritumoural progression area was done by using a supervised convolutional neural network. There was an overall accuracy of 92.6% in the training set and 78.5% in the validation set. Multimodal MR radiomics can demonstrate distinct characteristics in areas of potential progression on preoperative MRI.

show abstract

mentioning

confidence: 99%

A Neural Network Approach to Identify the Peritumoral Invasive Areas in Glioblastoma Patients by Using MR Radiomics

Yan

Hoorn

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…In summary, high CBV values of nonenhancing T2 hyperintense lesions seemed to be the most useful parameter for LR prediction. Kim et al also reported that CBV features from nonenhancing regions were useful for anticipating local recurrence, and even more signi cant if combined with franctional anisotrpy 20 . Our LR prediction model results determined that in the CBV maps, nonenhancing area were more competent than contrast-enhanced regions.…”

Section: Discussionmentioning

confidence: 99%

“…Though we reduced the over tting due to high-dimensionality by using dimension reduction, and data augmentation, the prediction model should be further improved using a larger dataset. In addition, radiomics data may be dependent on MR scanner or sequence settings, which was equalized in this study 20 . Further research with multicentered larger sample size with prospective design should be conducted to validate the generalizability of the developed model.…”

Section: Discussionmentioning

confidence: 99%

Prediction of Recurrence Patterns in Glioblastoma Using DSC-MRI Radiomics-Based Deep Learning

Ky¹,

Sw²,

Jh³

et al. 2020

Preprint

View full text Add to dashboard Cite

Glioblastoma remains the most devastating brain tumor despite optimal treatment, because of the high rate of recurrence. Distant recurrence has distinct genomic alterations compared to local recurrence, which requires different treatment planning both in clinical practice and trials. To date, perfusion-weighted MRI has revealed that perfusional characteristics of tumor are associated with prognosis. However, not much research has focused on recurrence patterns in glioblastoma: namely, local and distant recurrence. Here, we propose two different neural network models to predict the recurrence patterns in glioblastoma that utilizes high-dimensional radiomic profiles based on perfusion MRI: area under the curve (AUC) (95% confidence interval), 0.969 (0.903-1.000) for local recurrence; 0.864 (0.726-0.976) for distant recurrence for each patient in the validation set. This creates an opportunity to provide personalized medicine in contrast to studies investigating only group differences. Moreover, interpretable deep learning identified that salient radiomic features for each recurrence pattern are related to perfusional intratumoral heterogeneity. We also demonstrated that the combined salient radiomic features, or “radiomic risk score”, increased risk of recurrence/progression (hazard ratio, 1.61; p=0.03) in multivariate Cox regression on progression-free survival.

show abstract

“…In summary, high CBV values of nonenhancing T2 hyperintense lesions seemed to be the most useful parameter for LR prediction. Kim et al also reported that CBV features from nonenhancing regions were useful for anticipating local recurrence, and even more significant if combined with franctional anisotrpy 20 . Our LR prediction model results determined that in the CBV maps, nonenhancing area were more competent than contrast-enhanced regions.…”

Section: Discussionmentioning

confidence: 99%

“…Other studies have similarly showed that CBV features, associated with tumor aggressiveness, is valuable parameters for prediction of glioma patient survival and prognosis 17 – 19 . However, single parameter approach has been noted for limited ability in survival prediction 20 , 21 . Recently, radiomics approach, in combined with building prediction models with large number of parameters such as neural networks has enabled accurate prediction at an individual level 8 , 22 , though the developed models should be sufficiently validated.…”

Section: Discussionmentioning

confidence: 99%

Radiomics-based neural network predicts recurrence patterns in glioblastoma using dynamic susceptibility contrast-enhanced MRI

Shim

Chung

Jeong

et al. 2021

Sci Rep

View full text Add to dashboard Cite

Glioblastoma remains the most devastating brain tumor despite optimal treatment, because of the high rate of recurrence. Distant recurrence has distinct genomic alterations compared to local recurrence, which requires different treatment planning both in clinical practice and trials. To date, perfusion-weighted MRI has revealed that perfusional characteristics of tumor are associated with prognosis. However, not much research has focused on recurrence patterns in glioblastoma: namely, local and distant recurrence. Here, we propose two different neural network models to predict the recurrence patterns in glioblastoma that utilizes high-dimensional radiomic profiles based on perfusion MRI: area under the curve (AUC) (95% confidence interval), 0.969 (0.903–1.000) for local recurrence; 0.864 (0.726–0.976) for distant recurrence for each patient in the validation set. This creates an opportunity to provide personalized medicine in contrast to studies investigating only group differences. Moreover, interpretable deep learning identified that salient radiomic features for each recurrence pattern are related to perfusional intratumoral heterogeneity. We also demonstrated that the combined salient radiomic features, or “radiomic risk score”, increased risk of recurrence/progression (hazard ratio, 1.61; p = 0.03) in multivariate Cox regression on progression-free survival.

show abstract

Radiomics in peritumoral non-enhancing regions: fractional anisotropy and cerebral blood volume improve prediction of local progression and overall survival in patients with glioblastoma

Cited by 45 publications

References 45 publications

A Neural Network Approach to Identify the Peritumoral Invasive Areas in Glioblastoma Patients by Using MR Radiomics

A Neural Network Approach to Identify the Peritumoral Invasive Areas in Glioblastoma Patients by Using MR Radiomics

Prediction of Recurrence Patterns in Glioblastoma Using DSC-MRI Radiomics-Based Deep Learning

Radiomics-based neural network predicts recurrence patterns in glioblastoma using dynamic susceptibility contrast-enhanced MRI

Contact Info

Product

Resources

About