Dataset Information

Standardization of brain MR images across machines and protocols: bridging the gap for MRI-based radiomics.

ABSTRACT: Radiomics relies on the extraction of a wide variety of quantitative image-based features to provide decision support. Magnetic resonance imaging (MRI) contributes to the personalization of patient care but suffers from being highly dependent on acquisition and reconstruction parameters. Today, there are no guidelines regarding the optimal pre-processing of MR images in the context of radiomics, which is crucial for the generalization of published image-based signatures. This study aims to assess the impact of three different intensity normalization methods (Nyul, WhiteStripe, Z-Score) typically used in MRI together with two methods for intensity discretization (fixed bin size and fixed bin number). The impact of these methods was evaluated on first- and second-order radiomics features extracted from brain MRI, establishing a unified methodology for future radiomics studies. Two independent MRI datasets were used. The first one (DATASET1) included 20 institutional patients with WHO grade II and III gliomas who underwent post-contrast 3D axial T1-weighted (T1w-gd) and axial T2-weighted fluid attenuation inversion recovery (T2w-flair) sequences on two different MR devices (1.5 T and 3.0 T) with a 1-month delay. Jensen-Shannon divergence was used to compare pairs of intensity histograms before and after normalization. The stability of first-order and second-order features across the two acquisitions was analysed using the concordance correlation coefficient and the intra-class correlation coefficient. The second dataset (DATASET2) was extracted from the public TCIA database and included 108 patients with WHO grade II and III gliomas and 135 patients with WHO grade IV glioblastomas. The impact of normalization and discretization methods was evaluated based on a tumour grade classification task (balanced accuracy measurement) using five well-established machine learning algorithms. Intensity normalization highly improved the robustness of first-order features and the performances of subsequent classification models. For the T1w-gd sequence, the mean balanced accuracy for tumour grade classification was increased from 0.67 (95% CI 0.61-0.73) to 0.82 (95% CI 0.79-0.84, P?=?.006), 0.79 (95% CI 0.76-0.82, P?=?.021) and 0.82 (95% CI 0.80-0.85, P?=?.005), respectively, using the Nyul, WhiteStripe and Z-Score normalization methods compared to no normalization. The relative discretization makes unnecessary the use of intensity normalization for the second-order radiomics features. Even if the bin number for the discretization had a small impact on classification performances, a good compromise was obtained using the 32 bins considering both T1w-gd and T2w-flair sequences. No significant improvements in classification performances were observed using feature selection. A standardized pre-processing pipeline is proposed for the use of radiomics in MRI of brain tumours. For models based on first- and second-order features, we recommend normalizing images with the Z-Score method and adopting an absolute discretization approach. For second-order feature-based signatures, relative discretization can be used without prior normalization. In both cases, 32 bins for discretization are recommended. This study may pave the way for the multicentric development and validation of MR-based radiomics biomarkers.

SUBMITTER: Carre A

PROVIDER: S-EPMC7378556 | biostudies-literature | 2020 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Standardization of brain MR images across machines and protocols: bridging the gap for MRI-based radiomics.

Carré Alexandre A Klausner Guillaume G Edjlali Myriam M Lerousseau Marvin M Briend-Diop Jade J Sun Roger R Ammari Samy S Reuzé Sylvain S Alvarez Andres Emilie E Estienne Théo T Niyoteka Stéphane S Battistella Enzo E Vakalopoulou Maria M Dhermain Frédéric F Paragios Nikos N Deutsch Eric E Oppenheim Catherine C Pallud Johan J Robert Charlotte C

Scientific reports 20200723 1

Radiomics relies on the extraction of a wide variety of quantitative image-based features to provide decision support. Magnetic resonance imaging (MRI) contributes to the personalization of patient care but suffers from being highly dependent on acquisition and reconstruction parameters. Today, there are no guidelines regarding the optimal pre-processing of MR images in the context of radiomics, which is crucial for the generalization of published image-based signatures. This study aims to asses ...[more]

PMID: 32704007

Similar Datasets

Project description:BackgroundPortable low-field-strength magnetic resonance imaging (MRI) systems represent a promising alternative to traditional high-field-strength systems with the potential to make MR technology available at scale in low-resource settings. However, lower image quality and resolution may limit the research and clinical potential of these devices. We tested two super-resolution methods to enhance image quality in a low-field MR system and compared their correspondence with images acquired from a high-field system in a sample of young people.MethodsT1- and T2-weighted structural MR images were obtained from a low-field (64mT) Hyperfine and high-field (3T) Siemens system in N = 70 individuals (mean age = 20.39 years, range 9-26 years). We tested two super-resolution approaches to improve image correspondence between images acquired at high- and low-field: (1) processing via a convolutional neural network ('SynthSR'), and (2) multi-orientation image averaging. We extracted brain region volumes, cortical thickness, and cortical surface area estimates. We used Pearson correlations to test the correspondence between these measures, and Steiger Z tests to compare the difference in correspondence between standard imaging and super-resolution approaches.ResultsSingle pairs of T1- and T2-weighted images acquired at low field showed high correspondence to high-field-strength images for estimates of total intracranial volume, surface area cortical volume, subcortical volume, and total brain volume (r range = 0.60-0.88). Correspondence was lower for cerebral white matter volume (r = 0.32, p = 0.007, q = 0.009) and non-significant for mean cortical thickness (r = -0.05, p = 0.664, q = 0.664). Processing images with SynthSR yielded significant improvements in correspondence for total brain volume, white matter volume, total surface area, subcortical volume, cortical volume, and total intracranial volume (r range = 0.85-0.97), with the exception of global mean cortical thickness (r = 0.14). An alternative multi-orientation image averaging approach improved correspondence for cerebral white matter and total brain volume. Processing with SynthSR also significantly improved correspondence across widespread regions for estimates of cortical volume, surface area and subcortical volume, as well as within isolated prefrontal and temporal regions for estimates of cortical thickness.ConclusionApplying super-resolution approaches to low-field imaging improves regional brain volume and surface area accuracy in young people. Finer-scale brain measurements, such as cortical thickness, remain challenging with the limited resolution of low-field systems.

Project description:BackgroundIt is difficult to identify pancreatic ductal adenocarcinoma (PDAC) and mass-forming chronic pancreatitis (MFCP) lesions through conventional CT or MR examination. As an innovative image analysis method, radiomics may possess potential clinical value in identifying PDAC and MFCP. To develop and validate radiomics models derived from multiparametric MRI to distinguish pancreatic ductal adenocarcinoma (PDAC) and mass-forming chronic pancreatitis (MFCP) lesions.MethodsThis retrospective study included 119 patients from two independent institutions. Patients from one institution were used as the training cohort (51 patients with PDAC and 13 patients with MFCP), and patients from the other institution were used as the testing cohort (45 patients with PDAC and 10 patients with MFCP). All the patients had pathologically confirmed results, and preoperative MRI was performed. Four feature sets were extracted from T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), and the artery (A) and portal (P) phases of dynamic contrast-enhanced MRI, and the corresponding radiomics models were established. Several clinical characteristics were used to discriminate PDAC and MFCP lesions, and clinical model was established. The results of radiologists' evaluation were compared with pathology and radiomics models. Univariate analysis and the least absolute shrinkage and selection operator algorithm were performed for feature selection, and a support vector machine was used for classification. The receiver operating characteristic (ROC) curve was applied to assess the model discrimination.ResultsThe areas under the ROC curves (AUCs) for the T1WI, T2WI, A and, P and clinical models were 0.893, 0.911, 0.958, 0.997 and 0.516 in the primary cohort, and 0.882, 0.902, 0.920, 0.962 and 0.649 in the validation cohort, respectively. All radiomics models performed better than clinical model and radiologists' evaluation both in the training and testing cohorts by comparing the AUC of various models, all P<0.050. Good calibration was achieved.ConclusionsThe radiomics models based on multiparametric MRI have the potential ability to classify PDAC and MFCP lesions.

Dataset Information

Standardization of brain MR images across machines and protocols: bridging the gap for MRI-based radiomics.

Publications

Standardization of brain MR images across machines and protocols: bridging the gap for MRI-based radiomics.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets