Accurately classifying colorectal polyps, or differentiating malignant from benign ones, has a significant clinical impact on early detection and identifying optimal treatment of colorectal cancer. Convolution neural network (CNN) has shown great potential in recognizing different objects (e.g. human faces) from multiple slice (or color) images, a task similar to the polyp differentiation, given a large learning database. This study explores the potential of CNN learning from multiple slice (or feature) images to differentiate malignant from benign polyps from a relatively small database with pathological ground truth, including 32 malignant and 31 benign polyps represented by volumetric computed tomographic (CT) images. The feature image in this investigation is the gray-level co-occurrence matrix (GLCM). For each volumetric polyp, there are 13 GLCMs, computed from each of the 13 directions through the polyp volume. For comparison purpose, the CNN learning is also applied to the multi-slice CT images of the volumetric polyps. The comparison study is further extended to include Random Forest (RF) classification of the Haralick texture features (derived from the GLCMs). From the relatively small database, this study achieved scores of 0.91/0.93 (two-fold/leave-one-out evaluations) AUC (area under curve of the receiver operating characteristics) by using the CNN on the GLCMs, while the RF reached 0.84/0.86 AUC on the Haralick features and the CNN rendered 0.79/0.80 AUC on the multiple-slice CT images. The presented CNN learning from the GLCMs can relieve the challenge associated with relatively small database, improve the classification performance over the CNN on the raw CT images and the RF on the Haralick features, and have the potential to perform the clinical task of differentiating malignant from benign polyps with pathological ground truth.

BACKGROUND AND PURPOSE: The heterogeneity of glioblastoma contributes to the poor and variant prognosis. The aim of this retrospective study was to assess the glioblastoma heterogeneity with MR imaging textures and to evaluate its impact on survival time. MATERIALS AND METHODS: A total of 133 patients with primary glioblastoma who underwent postcontrast T1-weighted imaging (acquired before treatment) and whose data were filed with the survival times were selected from the Cancer Genome Atlas. On the basis of overall survival, the patients were divided into 2 groups: long-term (≥12 months, n = 67) and short-term (<12 months, n = 66) survival. To measure heterogeneity, we extracted 3 types of textures, co-occurrence matrix, run-length matrix, and histogram, reflecting local, regional, and global spatial variations, respectively. Then the support vector machine classification was used to determine how different texture types perform in differentiating the 2 groups, both alone and in combination. Finally, a recursive feature-elimination method was used to find an optimal feature subset with the best differentiation performance. RESULTS: When used alone, the co-occurrence matrix performed best, while all the features combined obtained the best survival stratification. According to feature selection and ranking, 43 top-ranked features were selected as the optimal subset. Among them, the top 10 features included 7 run-length matrix and 3 co-occurrence matrix features, in which all 6 regional run-length matrix features emphasizing high gray-levels ranked in the top 7. CONCLUSIONS: The results suggest that local and regional heterogeneity may play an important role in the survival stratification of patients with glioblastoma.

We present a fully automatic algorithm for brain magnetic resonance (MR) image segmentation. The three-dimensional (3D) volumetric MR dataset is first interpolated for an adequate local intensity vector on each voxel. Then a morphology dilation filter and region growing technique are applied to extract the region of brain volume, strapping away the skull, scalp and other tissues. The principal component analysis (PCA) is utilized to generate a series of feature vectors from the local vectors via the Karhunen-Loeve (K-L) transformation for those voxels within the extracted region. We choose those first few principal components that sum up to, at least, 90% percent of the total variance for optimizing the dimensions of the feature vectors. Then a modified self-adaptive online vector quantization algorithm is applied to these feature vectors for classification. The presented algorithm requires no prior knowledge of the data distribution except a maximum number of distinct groups for classification, which can be set based on anatomical knowledge. Numerical analysis of the algorithm and experimental tests on brain MR images are presented. Results demonstrate efficient, robust, and self-adaptive properties of the presented algorithm.

Most penalized maximum likelihood methods for tomographic image reconstruction based on Bayes’ law include a freely adjustable hyperparameter to balance the data fidelity term and the prior/penalty term for a specific noise–resolution tradeoff. The hyperparameter is determined empirically via a trial-and-error fashion in many applications, which then selects the optimal result from multiple iterative reconstructions. These penalized methods are not only time-consuming by their iterative nature, but also require manual adjustment. This study aims to investigate a theory-based strategy for Bayesian image reconstruction without a freely adjustable hyperparameter, to substantially save time and computational resources. The Bayesian image reconstruction problem is formulated by two probability density functions (PDFs), one for the data fidelity term and the other for the prior term. When formulating these PDFs, we introduce two parameters. While these two parameters ensure the PDFs completely describe the data and prior terms, they cannot be determined by the acquired data; thus, they are called complete but unobservable parameters. Estimating these two parameters becomes possible under the conditional expectation and maximization for the image reconstruction, given the acquired data and the PDFs. This leads to an iterative algorithm, which jointly estimates the two parameters and computes the to-be reconstructed image by maximizing a posteriori probability, denoted as joint-parameter-Bayes. In addition to the theoretical formulation, comprehensive simulation experiments are performed to analyze the stopping criterion of the iterative joint-parameter-Bayes method. Finally, given the data, an optimal reconstruction is obtained without any freely adjustable hyperparameter by satisfying the PDF condition for both the data likelihood and the prior probability, and by satisfying the stopping criterion. Moreover, the stability of joint-parameter-Bayes is investigated through factors such as initialization, the PDF specification, and renormalization in an iterative manner. Both phantom simulation and clinical patient data results show that joint-parameter-Bayes can provide comparable reconstructed image quality compared to the conventional methods, but with much less reconstruction time. To see the response of the algorithm to different types of noise, three common noise models are introduced to the simulation data, including white Gaussian noise to post-log sinogram data, Poisson-like signal-dependent noise to post-log sinogram data and Poisson noise to the pre-log transmission data. The experimental outcomes of the white Gaussian noise reveal that the two parameters estimated by the joint-parameter-Bayes method agree well with simulations. It is observed that the parameter introduced to satisfy the prior’s PDF is more sensitive to stopping the iteration process for all three noise models. A stability investigation showed that the initial image by filtered back projection is very robust. Clinical patient data demonstrated the effectiveness of the proposed joint-parameter-Bayes and stopping criterion.