Witnessed the development of deep learning, increasing number of studies try to build computer aided diagnosis systems for 3D volumetric medical data. However, as the annotations of 3D medical data are difficult to acquire, the number of annotated 3D medical images is often not enough to well train the deep learning networks. The self-supervised learning deeply exploiting the information of raw data is one of the potential solutions to loose the requirement of training data. In this paper, we propose a self-supervised learning framework for the volumetric medical images. A novel proxy task, i.e., Rubik's cube recovery, is formulated to pre-train 3D neural networks. The proxy task involves two operations, i.e., cube rearrangement and cube rotation, which enforce networks to learn translational and rotational invariant features from raw 3D data. Compared to the train-from-scratch strategy, fine-tuning from the pretrained network leads to a better accuracy on various tasks, e.g., brain hemorrhage classification and brain tumor segmentation. We show that our self-supervised learning approach can substantially boost the accuracies of 3D deep learning networks on the volumetric medical datasets without using extra data. To our best knowledge, this is the first work focusing on the self-supervised learning of 3D neural networks.

Computed tomography (CT) can provide a 3D view of the patient's internal organs, facilitating disease diagnosis, but it incurs more radiation dose to a patient and a CT scanner is much more cost prohibitive than an X-ray machine too. Traditional CT reconstruction methods require hundreds of X-ray projections through a full rotational scan of the body, which cannot be performed on a typical X-ray machine. In this work, we propose to reconstruct CT from two orthogonal X-rays using the generative adversarial network (GAN) framework. A specially designed generator network is exploited to increase data dimension from 2D (X-rays) to 3D (CT), which is not addressed in previous research of GAN. A novel feature fusion method is proposed to combine information from two X-rays. The mean squared error (MSE) loss and adversarial loss are combined to train the generator, resulting in a high-quality CT volume both visually and quantitatively. Extensive experiments on a publicly available chest CT dataset demonstrate the effectiveness of the proposed method. It could be a nice enhancement of a low-cost X-ray machine to provide physicians a CT-like 3D volume in several niche applications.

Pyrethroids are the major class of insecticides used for mosquito control. Excessive and improper use of insecticides, however, has resulted in pyrethroid resistance, which has become a major obstacle for mosquito control. The development of pyrethroid resistance is a complex process involving many genes, and information on post-transcription regulation of pyrethroid resistance is lacking. In this study, we extracted RNA from mosquitoes in various life stages (fourth-instar larvae, pupae, male and female adult mosquitoes) from deltamethrin-sensitive (DS) and resistant (DR) strains. Using Illumina sequencing, we obtained 13760296 and 12355472 reads for DS-strains and DR-strains, respectively. We identified 100 conserved miRNAs and 42 novel miRNAs derived from 21 miRNA precursors in Culex pipiens. After normalization, we identified 28 differentially expressed miRNAs between the two strains. Additionally, we found that cpp-miR-71 was significant down regulated in female adults from the DR-strain. Based on microinjection and CDC Bottle Bioassay data, we found that cpp-miR-71 may play a contributing role in deltamethrin resistance. The present study provides the firstly large-scale characterization of miRNAs in Culex pipiens and provides evidence of post-transcription regulation. The differentially expressed miRNAs between the two strains are expected to contribute to the development of pyrethroid resistance.

Esophageal cancer is one of the most common malignancies in the world. Squamous cell carcinoma accounts for approximately 90 % of esophageal cancer cases. Genetic and epigenetic changes have been found to accumulate during the development of various cancers, including esophageal squamous carcinoma (ESCC). Tobacco smoking and alcohol consumption are two major risk factors for ESCC, and both tobacco and alcohol were found to induce methylation changes in ESCC. Growing evidence demonstrates that aberrant epigenetic changes play important roles in the multiple-step processes of carcinogenesis and tumor progression. DNA methylation may occur in the key components of cancer-related signaling pathways. Aberrant DNA methylation affects genes involved in cell cycle, DNA damage repair, Wnt, TGF-β, and NF-κB signaling pathways, including P16, MGMT, SFRP2, DACH1, and ZNF382. Certain genes methylated in precursor lesions of the esophagus demonstrate that DNA methylation may serve as esophageal cancer early detection marker, such as methylation of HIN1, TFPI-2, DACH1, and SOX17. CHFR methylation is a late stage event in ESCC and is a sensitive marker for taxanes in human ESCC. FHIT methylation is associated with poor prognosis in ESCC. Aberrant DNA methylation changes may serve as diagnostic, prognostic, and chemo-sensitive markers. Characterization of the DNA methylome in ESCC will help to better understand its mechanisms and develop improved therapies.

Purpose Microvascular invasion (MVI) is a valuable predictor of survival in hepatocellular carcinoma (HCC) patients. This study developed predictive models using eXtreme Gradient Boosting (XGBoost) and deep learning based on CT images to predict MVI preoperatively. Methods In total, 405 patients were included. A total of 7302 radiomic features and 17 radiological features were extracted by a radiomics feature extraction package and radiologists, respectively. We developed a XGBoost model based on radiomics features, radiological features and clinical variables and a three-dimensional convolutional neural network (3D-CNN) to predict MVI status. Next, we compared the efficacy of the two models. Results Of the 405 patients, 220 (54.3%) were MVI positive, and 185 (45.7%) were MVI negative. The areas under the receiver operating characteristic curves (AUROCs) of the Radiomics-Radiological-Clinical (RRC) Model and 3D-CNN Model in the training set were 0.952 (95% confidence interval (CI) 0.923–0.973) and 0.980 (95% CI 0.959–0.993), respectively (p = 0.14). The AUROCs of the RRC Model and 3D-CNN Model in the validation set were 0.887 (95% CI 0.797–0.947) and 0.906 (95% CI 0.821–0.960), respectively (p = 0.83). Based on the MVI status predicted by the RRC and 3D-CNN Models, the mean recurrence-free survival (RFS) was significantly better in the predicted MVI-negative group than that in the predicted MVI-positive group (RRC Model: 69.95 vs. 24.80 months, p < 0.001; 3D-CNN Model: 64.06 vs. 31.05 months, p = 0.027). Conclusion The RRC Model and 3D-CNN models showed considerable efficacy in identifying MVI preoperatively. These machine learning models may facilitate decision-making in HCC treatment but requires further validation.

No abstract

Retinal artery/vein (A/V) classification plays a critical role in the clinical biomarker study of how various systemic and cardiovascular diseases affect the retinal vessels. Conventional methods of automated A/V classification are generally complicated and heavily depend on the accurate vessel segmentation. In this paper, we propose a multi-task deep neural network with spatial activation mechanism that is able to segment full retinal vessel, artery and vein simultaneously, without the pre-requirement of vessel segmentation. The input module of the network integrates the domain knowledge of widely used retinal preprocessing and vessel enhancement techniques. We specially customize the output block of the network with a spatial activation mechanism, which takes advantage of a relatively easier task of vessel segmentation and exploits it to boost the performance of A/V classification. In addition, deep supervision is introduced to the network to assist the low level layers to extract more semantic information. The proposed network achieves pixel-wise accuracy of 95.70% for vessel segmentation, and A/V classification accuracy of 94.50%, which is the state-of-the-art performance for both tasks on the AV-DRIVE dataset. Furthermore, we have also tested the model performance on INSPIRE-AVR dataset, which achieves a skeletal A/V classification accuracy of 91.6%.

Macrophages play a crucial role in tumorigenesis depending upon the phenotype of macrophages found in tumor microenvironments. To date, how the tumor microenvironment affects the phenotypes of macrophages is not yet fully understood. In this study, we constructed a NIH3T3/Src cell line stably overexpresses the Src protein and found that conditioned medium from this cell line was able to induce polarization towards the M2 phenotype in primary bone marrow-derived macrophages (BMDM) and Ana-1 macrophages. Further investigation revealed that IL-6 produced by NIH3T3/Src cells plays a key role in M2 polarization. During the development of colorectal cancer in C57BL/6J-ApcMin/+ mice, increased IL-6 secretion in the interstitial fluid of the colorectal tissues was observed. Furthermore, tumorigenesis in IL-6tm1Kopf mice treated with AOM-DSS, an IL-6 knockout mouse strain, was significantly inhibited compared with the control group, suggesting the important role of IL-6 in promoting tumorigenicity. Our findings identify the target molecules and proinflammatory cytokines responsible for promoting polarization towards the M2 phenotype in macrophages present in tumor microenvironment, which may be useful for the design of novel therapeutic strategies for colorectal cancer.