Search

Liver cancer is one of the most common malignant tumors. In Asia, liver cancer often develops on a background of cirrhosis caused by chronic hepatitis. The procedure of hepatitis, cirrhotic nodules, dysplastic nodules, and then hepatocellular carcinoma is the most common liver cancer evolutionary process. Judging the stage of hepatic nodules in the evolution process and taking intervention measures are critical for reducing the incidence of liver cancer. In this paper, a more accurate support vector machine (SVM) classification algorithm, LFOA-F-SVM, was proposed for radiomics to classify hepatic nodules from 120 patients into four categories based on dynamic enhanced magnetic resonance images. The algorithm uses radius-margin-based F-SVM, and combines the fruit fly optimization algorithm (FOA) of Levy flight (LF) strategy to optimize the parameters. To verify the effectiveness of the method, five UCI classification data sets (hearts, Parkinson’s disease, iris, wine and zoo) were added and compared with SVM, PSO-SVM, FOA-SVM, F-SVM. The results showed that LFOA-F-SVM has the highest classification accuracy in six data sets compared to the other methods. And in the hepatic nodules data set, the classification precision and recall are relatively high.

The left myocardium is the strongest cardiac muscle, representing the ability of heart to pump arterial blood throughout the body. Quantitative analysis of the contractive motion of left ventricle (LV) stands an important approach to diagnose cardiovascular diseases such as myocardial infarction. This paper utilized a biomechanical model describing the material of left myocardium to reconstruct the displacement field of LV. The mechanical model was used as an interpolation term, which was incorporated into the Bayesian estimation framework with the displacement field observed by cardiac cine magnetic resonance (CCMR) image, and the equation of displacement field was solved by finite element method. Functional parameters of LV were compared between the weak (46 cases) and normal (55 cases) ejection fraction groups of LV in the experiment. Significant differences in radial and circumferential displacement and velocity, strain and strain rate were observed between the two groups (p < 0.001), which proved that the method could effectively distinguish whether the motion of LV is normal or not. The experimental results are also highly correlated with the functional parameters of LV measured by CVI software, indicating that the method will facilitate the diagnosis of cardiovascular diseases.

In this paper, a new method for fine brain functional parcellation based on resting-state functional magnetic resonance imaging (rs-fMRI) data was proposed. The method combines the t-distribution stochastic neighbor embedding (t-SNE) and automatic spectral clustering (ASC) algorithms. First, correlation analyses are conducted between the time courses of the brain region to be parcellated and the whole brain. Second, t-SNE is used to extract the high-dimensional functional connectivity patterns. Last, the number of clusters is automatically determined by the ASC algorithm, and to divide the brain region of interest to generate the fine brain subregions. The results of simulated seed regions proved that the method proposed had higher accuracy than the commonly-used spectral clustering and spectral clustering with principal component analysis. Moreover, the method was successfully applied to parcellate the parahippocampal gyrus into 3 functional subregions in the left and right hemispheres. In conclusion, the algorithm combining t-SNE and ASC is an effective method for fine brain functional parcellation and construction of functional brain atlas.