Multimodal MRI Image Decision Fusion-Based Network for Glioma Classification

Abstract

PurposeGlioma is the most common primary brain tumor, with varying degrees of aggressiveness and prognosis. Accurate glioma classification is very important for treatment planning and prognosis prediction. The main purpose of this study is to design a novel effective algorithm for further improving the performance of glioma subtype classification using multimodal MRI images.MethodMRI images of four modalities for 221 glioma patients were collected from Computational Precision Medicine: Radiology-Pathology 2020 challenge, including T1, T2, T1ce, and fluid-attenuated inversion recovery (FLAIR) MRI images, to classify astrocytoma, oligodendroglioma, and glioblastoma. We proposed a multimodal MRI image decision fusion-based network for improving the glioma classification accuracy. First, the MRI images of each modality were input into a pre-trained tumor segmentation model to delineate the regions of tumor lesions. Then, the whole tumor regions were centrally clipped from original MRI images followed by max–min normalization. Subsequently, a deep learning-based network was designed based on a unified DenseNet structure, which extracts features through a series of dense blocks. After that, two fully connected layers were used to map the features into three glioma subtypes. During the training stage, we used the images of each modality after tumor segmentation to train the network to obtain its best accuracy on our testing set. During the inferring stage, a linear weighted module based on a decision fusion strategy was applied to assemble the predicted probabilities of the pre-trained models obtained in the training stage. Finally, the performance of our method was evaluated in terms of accuracy, area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), etc.ResultsThe proposed method achieved an accuracy of 0.878, an AUC of 0.902, a sensitivity of 0.772, a specificity of 0.930, a PPV of 0.862, an NPV of 0.949, and a Cohen’s Kappa of 0.773, which showed a significantly higher performance than existing state-of-the-art methods.ConclusionCompared with current studies, this study demonstrated the effectiveness and superiority in the overall performance of our proposed multimodal MRI image decision fusion-based network method for glioma subtype classification, which would be of enormous potential value in clinical practice.

Segmentation Algorithm of Magnetic Resonance Imaging Glioma under Fully Convolutional Densely Connected Convolutional Networks

“…Wang et al [ 38 ] optimized it based on the DenseNet algorithm, although its recognition and segmentation performance in the complete tumor region and core tumor region was improved, its DSC in the enhanced region was still low. The algorithm established in the literature [ 39 ] has a DSC value of 0.75 in the segmentation of glioma MRI images, and the DSC value of the CGSSNet algorithm in this paper in the segmentation of glioma MRI images was significantly higher than that in this work. It showed that the CGSSNet algorithm established in this work can be used to extract multiscale semantic information and improve the recognition and segmentation capabilities of glioma MRI images.…”

Artificial intelligence in multiparametric magnetic resonance imaging: A review

“…DL was first utilized to replace the classifier in hand‐crafted radiomics‐based methods 153 . With further technique development, end‐to‐end DCNNs have been developed to classify gliomas directly from images without the troublesome engineered image feature extraction process 154,162,163,169,170 …”

“…153 With further technique development,end-to-end DCNNs have been developed to classify gliomas directly from images without the troublesome engineered image feature extraction process. 154,162,163,169,170 In addition to glioma grading, accurate classification of glioma subtypes has also been extensively investigated utilizing machine learning-based methods. Prediction of glioma subtypes related to a single genetic alteration (e.g., the O 6 -methylguanine methyltransferase methylation status, 140,141 the isocitrate dehydrogenase mutation status, 142,[145][146][147][150][151][152]155,165,171 the nuclear alpha thalassemia/mental retardation syndrome X-linked expression, 151 the 1p/19q codelection status, 156,157 the telomerase reverse transcriptase mutation status, 158,159 and the vascular endothelial growth factor expression 171 ) has been frequently considered.…”

Role of Transmembrane Water Exchange in Glioma Invasion/Migration: In Vivo Preclinical Study by Relaxometry at Very Low Magnetic Field

“…FLAIR and DTI have both improved the assessment of the spatial extension of glioma, and new related biomarkers are still being proposed that correlate to disease [ 9 ] or glioma classification [ 10 ], yet the mechanisms responsible for these contrasts sources are not fully understood. Neuronavigation systems for preoperative and intraoperative MRI were also proposed to improve glioma invasion detection for efficient surgery, but both fail to delineate tumor margins.…”