Diagnosis of brain tumor gliomas is a challenging task in medical image analysis due to its complexity, the less regularity of tumor structures, and the diversity of tissue textures and shapes. Semantic segmentation approaches using deep learning have consistently outperformed the previous methods in this challenging task. However, deep learning is insufficient to provide the required local features related to tissue texture changes due to tumor growth. This paper designs a hybrid method arising from this need, which incorporates machine-learned and hand-crafted features. A semantic segmentation network (SegNet) is used to generate the machine-learned features, while the grey-level co-occurrence matrix (GLCM)-based texture features construct the hand-crafted features. In addition, the proposed approach only takes the region of interest (ROI), which represents the extension of the complete tumor structure, as input, and suppresses the intensity of other irrelevant area. A decision tree (DT) is used to classify the pixels of ROI MRI images into different parts of tumors, i.e. edema, necrosis and enhanced tumor. The method was evaluated on BRATS 2017 dataset. The results demonstrate that the proposed model provides promising segmentation in brain tumor structure. The F-measures for automatic brain tumor segmentation against ground truth are 0.98, 0.75 and 0.69 for whole tumor, core and enhanced tumor, respectively.

The objective of this study is to expound the CT features of COVID-19 patients whose throat swab samples were negative for two consecutive nucleic acid tests after treatment. We retrospectively reviewed 46 COVID-19 patients with two consecutive negative RT-PCR tests after treatment. The cases were divided into moderate group and severe/critical group according to disease severity. Clinical and CT scanning data were collected. CT signs of pulmonary lesions and the score of lung involvement were expounded. Thirty-nine moderate cases and seven severe/critical cases were included. Residual pulmonary lesions were visible in CT images. Moderate patients showed peripheral lesions while severe/critical cases exhibited both central and peripheral lesions with all lobes involvement. Mixed ground glass opacity (GGO) and pulmonary consolidation were noted. A larger proportion of severe patients showed reticular pulmonary interstitium thickening. Air bronchogram, pleural effusion, vascular enlargement, bronchial wall thickening, bronchiectasis, pleural thickening and pleural adhesion were more frequently observed in severe/critical group. The severe/critical group showed higher CT score. Pulmonary lesions persisted even after twice consecutive negative nucleic acid tests. We strongly recommended regular follow-up of CT scans after nucleic acid tests conversion. Evaluation of complete remission should base on chest CT.

Brain tumour tissue segmentation is essential for clinical decision making. While manual segmentation is time consuming, tedious, and subjective, it is very challenging to develop automatic segmentation methods. Deep learning with convolutional neural network (CNN) architecture has consistently outperformed previous methods on such challenging tasks. However, the local dependencies of pixel classes cannot be fully reflected in the CNN models. In contrast, hand-crafted features such as histogram-based texture features provide robust feature descriptors of local pixel dependencies. In this paper, a classification-based method for automatic brain tumour tissue segmentation is proposed using combined CNN-based and hand-crafted features. The CIFAR network is modified to extract CNN-based features, and histogram-based texture features are fused to compensate the limitation in the CIFAR network. These features together with the pixel intensities of the original MRI images are sent to a decision tree for classifying the MRI image voxels into different types of tumour tissues. The method is evaluated on the BraTS 2017 dataset. Experiments show that the proposed method produces promising segmentation results.

Abstract Background: Early recognition of patients who are prone to develop severe or critical COVID-19 pneumonia may improve its management and modify its outcome by better treatment regimens.Objective: To investigate Computed Tomography (CT)-based early markers of COVID-19 progression. Material and methods: Thirty-two COVID-19 patients were retrospectively enrolled, including 23 moderate cases who remained moderate until being cured and discharged from hospital, and 9 progressive cases who started with moderate pneumonia and later progressed to severe or critical pneumonia. Clinical and CT data in the early course of disease were reviewed and compared. The association between early features and progression of COVID-19 were analyzed. Results: The multivariate logistic regression analysis revealed that the progression of COVID-19 were significantly associated with older age and higher CT score of the right lower lobe in the early CT images. After adjustment for relevant covariates, the relationship between CT score of the right lower lobe and COVID-19 progression persisted (odds ratio 19.28, 95% confidence interval 1.07 - 348.34).Conclusions: Higher CT Score of the right lower lobe in early images and older age appeared to be promising biomarkers for early prediction of COVID-19 prognosis. Extensive involvement of the right lower lobe at the early stage may suggest a high risk of progression.