Chinese and global burdens of gastric cancer from 1990 to 2019.

BACKGROUND: Gastric cancer is a common cancer in China. This project investigated the disease burden of gastric cancer from 1990 to 2019 in China and globally. METHODS: The global age-standardized rates (ASRs) were extracted from the Global Burden of Disease. Moreover, the estimated annual percentage changes (eAPCs) in the ASRs of incidence (ASIR), mortality (ASMR), and disability-adjusted life-years (DALYs) were calculated to determine the trends by countries and regions. RESULTS: In China, the ASIR declined from 37.56 to 30.64 per 100,000 and the ASMR declined from 37.73 to 21.72 per 100,000. The global ASIR decreased from 22.44 to 15.59 and the ASMR declined from 20.48 to 11.88 per 100,000 persons from 1990 to 2019. The ASIR was the lowest in Malawi (3.28 per 100,000) and the highest in Mongolia (43.7 per 100,000), whereas the ASMR was the lowest in the United States of America (3.40 per 100,000) and the highest in Mongolia (40.04 per 100,000) in 2019. The incidence of early-onset gastric cancer increased in China. The DALYs attributed to gastric cancer presented a slight decrease during the period. China had a higher mortality/incidence ratio (0.845) and 5-year prevalence (27.6/100,000) than most developed countries. CONCLUSION: China presented a steady decline in the incidence and mortality rates for gastric cancer. The global ASIR, ASMR, and DALYs showed a slight rise decrease. Different patterns of gastric cancer rates and temporal trends have been identified in different geographical regions, indicating that specific strategies are needed to prevent the increase in some countries.

A two-step automated quality assessment for liver MR images based on convolutional neural network.

PURPOSE: To propose an automatic approach based on a convolutional neural network (CNN) to evaluate the quality of T2-weighted liver magnetic resonance (MR) images as nondiagnostic (ND) or diagnostic (D). MATERIALS AND METHODS: We included 150 T2-weighted liver MR imaging examinations in this retrospective study. Each slice of liver image was annotated with a label D or ND by two radiologists with seven and six years of experience, respectively. Additionally, the radiologists segmented the liver region manually as the ground truth for liver segmentation. A CNN was trained to segment the liver region and another CNN was used to classify the qualities of patches extracted from the liver region. The quality of an image was obtained from the percentage of nondiagnostic patches in all liver patches in the image. Treating nondiagnostic as positive, the accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), area under the receiver operating characteristic curve (AUC), and confusion matrix were used to evaluate our model. A Mann-Whitney U test was performed with the statistical significance set at 0.05. RESULTS: Our model achieved good performance with an accuracy of 88.3 %, sensitivity of 86.0 %, specificity of 89.4 %, PPV of 78.6 %, NPV of 93.4 %, and AUC of 0.911 (95 % confidence interval: 0.882-0.939, p < 0.05). The confusion matrix of our model indicated good concordance with that of the radiologists. CONCLUSIONS: The proposed two-step patch-based model achieved excellent performance when assessing the quality of liver MR images.

Assessment of imprinting- and genetic variation-dependent monoallelic expression using reciprocal allele descendants between human family trios.

Genomic imprinting is an important epigenetic process that silences one of the parentally-inherited alleles of a gene and thereby exhibits allelic-specific expression (ASE). Detection of human imprinting events is hampered by the infeasibility of the reciprocal mating system in humans and the removal of ASE events arising from non-imprinting factors. Here, we describe a pipeline with the pattern of reciprocal allele descendants (RADs) through genotyping and transcriptome sequencing data across independent parent-offspring trios to discriminate between varied types of ASE (e.g., imprinting, genetic variation-dependent ASE, and random monoallelic expression (RME)). We show that the vast majority of ASE events are due to sequence-dependent genetic variant, which are evolutionarily conserved and may themselves play a cis-regulatory role. Particularly, 74% of non-RAD ASE events, even though they exhibit ASE biases toward the same parentally-inherited allele across different individuals, are derived from genetic variation but not imprinting. We further show that the RME effect may affect the effectiveness of the population-based method for detecting imprinting events and our pipeline can help to distinguish between these two ASE types. Taken together, this study provides a good indicator for categorization of different types of ASE, opening up this widespread and complex mechanism for comprehensive characterization.