Search

Medical image segmentation and registration have vital roles in computer-assisted diagnosis procedures, challenging tasks suffering from various limitations and artifacts inside images. Recently, deep learning techniques accomplish these two tasks and achieve outstanding performances. However, most deep learning-based methods overlook the potential correlation between each other. In this paper, an unsupervised probabilistic model named U-RSNet is proposed to realize concurrent medical image registration and segmentation in one framework. Specifically, the unsupervised segmentation branch is derived from Bayesian inference. The prior warped atlas for segmentation can be obtained by deforming a known probabilistic atlas by the corresponding invertible deformation field with a well-behaved diffeomorphic guarantee, which can perfectly integrate these two tasks to form a complete intelligent prediction system. In this case, the segmentation performance could be largely improved based on the warped probabilistic atlas obtained from the registration branch. Experiments on human brain 3D magnetic resonance images have demonstrated the effectiveness of our approach. We trained and validated U-RSNet with 1000 images and tested its performances on four public datasets. We showed our method successfully realized concurrent segmentation and registration and yielded better segmentation results than a separately trained network. [ABSTRACT FROM AUTHOR]

[Display omitted] • Propose an FCN architecture using octave partial convolution and skip connection with scSE (skip-scSE). • Design a novel decoder efficiently using a skip-scSE module. • Design a multi-modal MRI to single MRI modality synthesizing model using proposed FCN as a generator and encoder–decoder discriminator using the cGAN learning approach. • Generate missing modalities for the TCGA dataset using the available modality (or modalities) and predict segmentation using both available and synthesized modalities. • Design a radiogenomic model by combining radiomic features such as intensity, shape, and geometry with gene expression profiling data. The accurate prognosis of glioblastoma multiforme (GBM) plays an essential role in planning correlated surgeries and treatments. The conventional models of survival prediction rely on radiomic features using magnetic resonance imaging (MRI). In this paper, we propose a radiogenomic overall survival (OS) prediction approach by incorporating gene expression data with radiomic features such as shape, geometry, and clinical information. We exploit TCGA (The Cancer Genomic Atlas) dataset and synthesize the missing MRI modalities using a fully convolutional network (FCN) in a conditional generative adversarial network (cGAN). Meanwhile, the same FCN architecture enables the tumor segmentation from the available and the synthesized MRI modalities. The proposed FCN architecture comprises octave convolution (OctConv) and a novel decoder, with skip connections in spatial and channel squeeze & excitation (skip-scSE) block. The OctConv can process low and high-frequency features individually and improve model efficiency by reducing channel-wise redundancy. Skip-scSE applies spatial and channel-wise excitation to signify the essential features and reduces the sparsity in deeper layers learning parameters using skip connections. The proposed approaches are evaluated by comparative experiments with state-of-the-art models in synthesis, segmentation, and overall survival (OS) prediction. We observe that adding missing MRI modality improves the segmentation prediction, and expression levels of gene markers have a high contribution in the GBM prognosis prediction, and fused radiogenomic features boost the OS estimation. [ABSTRACT FROM AUTHOR]

Glioblastoma multiforme (GBM) are aggressive brain tumors, which lead to poor overall survival (OS) of patients. OS prediction of GBM patients provides useful information for surgical and treatment planning. Radiomics research attempts at predicting disease prognosis, thus providing beneficial information for personalized treatment from a variety of imaging features extracted from multiple MR images. In this study, MR image derived texture features, tumor shape and volumetric features, and patient age were obtained for 163 patients. OS group prediction was performed for both 2-class (short and long) and 3-class (short, medium and long) survival groups. Support vector machine classification based recursive feature elimination method was used to perform feature selection. The performance of the classification model was assessed using 5-fold cross-validation. The 2-class and 3-class OS group prediction accuracy obtained were 98.7% and 88.95% respectively. The shape features used in this work have been evaluated for OS prediction of GBM patients for the first time. The feature selection and prediction scheme implemented in this study yielded high accuracy for both 2-class and 3-class OS group predictions. This study was performed using routinely acquired MR images for GBM patients, thus making the translation of this work into a clinical setup convenient. [ABSTRACT FROM AUTHOR]